This article was downloaded by: [New York University]

On: 15 February 2015, At: 06:16

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

Annals of the Association of

American Geographers
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/raag20

Environment, Village and

City: A Genetic Approach to
Urban Geography; with Some
Reference to Possibilism
Griffith Taylor
Published online: 05 Mar 2009.

To cite this article: Griffith Taylor (1942) Environment, Village and City: A Genetic
Approach to Urban Geography; with Some Reference to Possibilism, Annals of the
Association of American Geographers, 32:1, 1-67

To link to this article: http://dx.doi.org/10.1080/00045604209357229

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.

This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
 sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [New York University] at 06:16 15 February 2015
 ANNALS
of the

Association of American Geographers

Volume XXXII MARCH, 1942 No. 1
_ _ ~
~- -~~ ~- -

Environment, Village and City

Downloaded by [New York University] at 06:16 15 February 2015

A Genetic Approach to Urban Geography;

with Some Reference to Possibilisni *
G R I F F I T H TAYLOR

INTRODUCTION

This is an outstanding year because it is marked by the complete econoiii‘c

and military alliance of the United States and the British Empire. They are
united in their deterinination not only to defeat enemy aggression, but also
to root out evil developments based on what is false in the Totalitarian soci-
ology. The international crisis makes me, as a Britisher, feel the more
honoured to be chosen to present the annual address to this Association of
Aiiiericaii Geographers.
The geographer studies environment. H e believes that it is a vital factor
in deterniining human progress. Some few of us who are willing to pro-
claim ourselves as, in some measure, “tarred with the determinist brush,”
think that this has been true right through man’s history. But the detailed
study of environment, with the production of accurate iiiaps and isopleth
charts, is a comparatively recent phase of research. Hence some of our
sister disciplines, such as history and anthropology, are not yet willing to
give much time to its study, or indeed to grant the necessity for such study.’
The present writer has devoted a good deal of attention to the way in
* Presidential address, delivered before the Association at New York, N. Y.,
December, 1941.
1 Historical GcograPhy of Englaad, edited by H. C. Darby (Cambridge, 1936) is
a fine example of the geographical approach to history.
1
 2 TAY LOR-GE NET1C URBAN GEOGRAP 13Y [Mar.

which enviroiiiiient, aiid especially changing environment, has affected major

human wzigratims. As regards their earliest phases this research has great
hearing on the fuiidanientals of anthropology. I have ventured to produce
a theory of racial evolution and classification which is making a good many
coiiverts among professional anthropologists. I n the later phases the study
of iiioderii migrations is of course linked up M it11 the development of pioneer
lands.’
If we now turn to the siiialler and perhaps more complex groups of nian
known as ~ i a t i o manother
, set of problems arises, which may be included in
the term Cultural Geography. The charting of national distributions and
their relation to environment have given rise to a field of research midway
Downloaded by [New York University] at 06:16 15 February 2015

between general geography and history, which the writer has found of great
interest.?
Finall) the still siirtrllo- examples oi hunian association are offered for our
attention i t i the form of cities, tonns, and villages. They are closely
anchored C t J the environment, and this relation has been engaging iiiy interest
since I first became associated with tlie original surveys for the projected
capital of h s t r a l i a , some thirty years ago. I believe that geographical tech-
nique will enable us to analyse the data connected with smaller human ag-
glomerations, and to develop Urban Geography as an orclerly department of
our discipline. This should be of service to the sociologist and town planner ;
just as, I firiiily believe, our techniques hare already helped tlie aiithropolo-
gist and historian.
Passing now to tlie third term in tlie title of my address. some explanation
may seem to be necessary for introducing the determinism-possibilisiii contro-
versy into this study. I t is precisely because I believe that man has more
“choice” of procedure as he is concerned with the details of geographic study,
that I am making the attempt to discriminate between the two controls (en-
vironmental aiid human) in this approach to Urban Geography. I think
that iiiy objection to the possibilist is not because lie s e e m often to be mainly
011 the lookout to see where nian has exercised his powers over Nature, but
because lie rather tends to ignore the rastly greater importance of environ-
mental control in most parts of the world. I have heard eminent geogra-
phers apologise for using the teriii “environmental control” ; aiid this, I
confess, annoys me.
I can make my point clearer perhaps by two exaiiiples drawn from lands
which I have studied in detail : Australia atid Antarctica. I would say that
the most important geographical problem in Australia is to try to picture its
population-pattern a century hence ; and then to develop Australia in that
2 Envivoizmcizt, Race, and Miqration, Griffith Taylor. Chicago, 1937.
3 Lr‘nviroizmc~rtttrrrd Nutioiz, Griffith Taylor. Chicago, 1936.
 19421 D E T E R X I N I S M VS. POSSIBILISM 3

direction. Tlie extreme possibilist says that such a project is quite visioii-
ary, because iiiaii’s plans and tools will change so greatly in tlie interval.
I have studied developments in other, older parts of tlie world, such as
Europe and Asia, aiid to me it seems clear that the general population-
pattern in the past has been determined almost exclusively by the environ-
ment ; and so it will be in the newer lands.
111Antarctica I have no doubt that a few way-stations for aviators, and
a few resorts for tourists, will develop there in the next century. These the
possibilist will eiiqhasize, as developing in spite of environment. But the
deteriiiinist will say that such trivial details, in an icy area of five million
square miles, are not worth mentioning in an attempt to evaluate enviroii-
Downloaded by [New York University] at 06:16 15 February 2015

mental versus human control.

Now these same principles hold in other better-endowed lands. The
progress of settleinent in North America is in the main essentials clue to
environmental control ; and we ouglit (as the chief interpreters of environ-
ment) to stress this fact. Where does the possibilist come i n ? As I see it,
lie is the short-range geographer, whereas tlie environmentalist takes a longer
view. Nature needs centuries for her influences to work properly. In most
major geographical problems there is one best way to do things, and of course
a dozen less satisfactory ways wliich man can “choose.” Usually iiiaii (and
in this he is, in a sense, supported by the possibilist) endeavours to proceed
by the long-established method of trial and error. H e chooses one possible
way and then another and finally adopts the one best suited to tlie environ-
ment. It is true that at times it is difficult to say which is the “best way,” but
the environmentalist is always debating the matter.
W e may use the sugar industry in northeast Australia as an illustration
of tlie difference in views. The possibilist points out that inaii has chosen
to grow sugar in this region, even though it can be purchased for half tlie
price in the foreign market. The long-range environmentalist would say
that Australia has not yet arrived at the stage where sugar should be grown
in her tropics. Australia is a pioneer country with a small population ; aiid
should at present concentrate on industries which require less costly labour.
Even if lie grants that tlie White Australia policy is worth some sacrifice,
it is the geographer’s duty, as I see it, to point out wherein man is departing
from the logical ordcl- of material development.

We see here why most economists and technologists dislike the

determinist idea. They are “short-range’’ developers ; and none the
less valuable teachers because that is their province. But the geogra-
pher, in my opinion, should look much farther ahead. H e should try
to find the best plan among those possible. Exaggerating somewhat,
I feel that man’s part in tlie programme of a country’s evolution is not
unlike that of a traffic policeman. In nine-tenths of tlie world-but
 4 TAYLOR-GENETIC URBAN GEOGRAPHY [Mar.

not so definitely iii the much restricted better-endowed parts of the

wo1-Id-a “Stop-and-Go Deterniiiiisni” seeins the logical ideology for
the geographer. Nature determines the route, iiian can only control
the rate of p r o g ~ e s s . ~
Let us now return to the topics which form the iiiaiii purpose of this
address. Though Urban Geography is a rather new development, there is
already a considerable literature on the subject. However I propose in this
address to discuss certain aspects of the growth of villages. towns, and cities
which I have investigated during iiiy rather wide experience as a \vaiidering
geographer.
Downloaded by [New York University] at 06:16 15 February 2015

It is difficult to decide where the geographical observer should set his

professional limits. The choice of the site of the young settlement is largely
controlled hy the environment. So also is its subsequent growth, excluding
perhaps the pattern of the streets. The form aiitl construction of the houses
depend largely on the environiiieiit (e.g., the climate) and on the materials
available. But admittedly many of the aspects of a town are clue to human
choice, and seem to have little to do with the local enviroiiment. I n these
aspects this geographer has less interest. Sometimes, indeed, the city seems
to develop in defiance of the enviroiiment; aiid this has led to waste and
unnecessary expense, as I hope to show later.
Several aspects of town development have riot been as fully discussed as
they might be by geographers. These, as well as some more familiar prob-
lems, I hope to lay before you in somewhat the following order :
a. Some variations of towns with latitude.
11. The city iiiade to plan, c.y., Canberra.
c. Tlie city as controlled by geological structure, e.g., Sydney.
d. Stages in the development of a city, e.g., Toronto.
e. The “Zones and Strata” concept as applied to towns ( P o r t Credit &
Whitby).
f . The classification of towns, and their description hy formulae.
g. Tlie bearing of this address 011 Deteriiiiiiisni I ’ ~ Y S I L . SPossibilism.
SI-CTIOS A
\

SEVEN EXAIZFLES O F S E T T L E M E N T S F R O M T Y P I C A L LATITUDES

In this address I do not hope to do more than draw attention to the vari-
ous aspects of human clusters which should he of interest in a genuine geo-
graphical approach. As a parallel I may cite the hotanist, who does not
need to know every unimportant detail about a plant i n order to classify it,
aiid to place it in its proper category. O r to come nearer home, Kiippeii has
produced a workable classification of the innumerable variations in the
4 Presidential Address to the British Geographers (B.A.A.S.), Griffith Taylor.

Cambridge, 1938.
 19421 SEVEN EXAMPLES O F SETTLEMENTS 5

world’s climates by concentrating on tlie main eleiuents of tlie problem. W e

have first to answer tlie question, what are tlie main elements in the growth
of a huniaii cluster, whether i n its early or its late stages?
I have therefore thought that it would be of interest to describe briefly
a number of towns and settlements which I have investigated in very differ-
ent parts of the world. Plans of each are illustrated, and some indications
of the surrounding region and of its influence on the agglomeration is given.
In several recent memoirs I have adopted tho “traverse” plan of study.
Thus in the northern Sahara I described3 the changes as one jouriieyed from
the wet coast to the southern desert. Later, in studying the main principles
of tlie Australian settlement in my recent I made a similar study and
investigated conditions at e p a l distances across the continent. As regards
Downloaded by [New York University] at 06:16 15 February 2015

the present study dealing with towiis i n general, we all know that latitude
is the chief variahle which affects settlement. Heiice the first part of this
address considers seven settlenients on a traverse from the Poles to the
Equator approximately 10” of latitude apart, each in a different continent.
Furthermore it will he of interest to try and decide the respective impor-
tances of the environment and the human factor in these type examples. NO
accurate evaluation is possible, but the writer (who is somewhat of a de-
terminist in the broader fields of racial and national distribution) is prepared
to admit that the human factor becomes of increasingly greater iiiiportance
as we consider smaller and smaller human agglomerations in our research.
The following table shows some of the main characters of the seven
samples chosen.
-~
-~ ~~~ ~
-~- __-
~

~ ~
~-
~~~~~ ~~~~~- ~

~~ ~ _ -~_ _ ~~ ~ ~~~~

~~~
~~ ~~

Popu-
Continent Place lation
-~ - _ ~--
Antarctica Cape Evans 25 Ice cap
Europe Bergen 100,000 Cool temper-
3 1 57 N. America McMurray
ate port
500 Taiga village
1 Fur,
Fish, etc.
etc.
Cfh
Dfc
4 ’ 40 Asia ’ Pekin 1,600,000 Oriental Political Dwa
5 36 Africa Biskra 4,000 Desert Dates; 1
6
7 i
22
11
Australia
S. America
Urandanj i
Santa
50 Semi-desert 1 military EWh
Cattle BShw
,
Marta 30,000 Tropical port , Danaiias Xw’i

Since the cliiiiatic data of these clusters are of importance in the picture,
I have added a hythergrapli chart (Fig. 1). I t shows that most o€ the chief
environments of tlie world are represented. I have also inserted a “C(11nfort
“Sea to Sahara,” Geog. REZ~., 29 (April, 1939).
Australia: A Sitidy of U’ar111 l:ii21ii o i i i i z m i s aiid Tlieir 6f f r c t on British Setflc-
niciit. New York, 1940.
 6 TAYLOR-GENETIC URBAN GEOGRAPH’I [Mas.
Downloaded by [New York University] at 06:16 15 February 2015

FIG.1.-Hythergraphs of the places described in Section A,, to show the wide-

spread character of the discussion. Koppen formulae are added. The “Comfort
Frame” includes the climates which are generally comfortable. Inset is a graph which
compares the standards of living in three regions (Algeria, Spain. and South Australia),
which all have the same environments.

Frame,” xvliicli tentatively encloses the climates which most Anglo-Ameri-

cans \vould call “comfortable.” I do not however propose to dwell in any
detail 011 climatic controls.

1. Cape Ezralis, a settleiiiciif 01.1 the Afitnrrtic ire cap

W e may learn a good deal as to the role of an adverse environinent in
the developmeiit of a settlement, by considering the first attempts to live in
Antarctica. No doubt one of the little coal towns in Spitsbergeii (such as
Longyear City) would be more inforiuative, but the sole experience of the
writer has been gained 011 the west coast of Ross Islalid in latitude 77” 30’
South.
The first question one should ask about any settleiiient i s “Why did inan
desire to establish himself there?” 111the majority of cases the answer is
 194.21 CAPE EVANS, ANTARCTICA 7
“Because he hoped to exploit Nature’s resources in the vicinity.” I n a very
remote degree this is true of the small settlements on Ross Island, of which
there are three (Fig. 2). Coal or metals may yet be found in rather inac-
Downloaded by [New York University] at 06:16 15 February 2015

FIG.2 (left).--A birds eye view of the west coast of Ross Island in the Antarctic
to show the settlements made by the three expeditions (1902, 1907, and 1910) where
the environment permitted. Dates of the ice freezing and breaking away are added.
FIG.2 (right).-Plan of the settlement at Cape Evans, Ross Island.

cessible areas in Antarctica, though there is little prospect of immediate use.

The major attraction was that hereabouts open water is found nearest
the South Pole. Hence Ross Island in the Ross Sea offered the best site for
exploration, and for extending in this unique environment our knowledge
of meteorology, magnetism, glacial erosion, geology, plant and animal ecol-
ogy, etc.
Secondly, why was Cape Evans chosen rather than any other site? Here
we have one of the best examples of environmental control known to the
writer. Every explorer wished to land his stores at the most southern spot
available, as e a d y in the suiiinier season as possible. Hence the huts were
invariably erected at the edge of the open sea, when the boats reached Ross
 S TAYLOR-G E N E T I C LIRRAK G E O G R A P H Y [Mar.

Island early in January. Thus in 1902 Scott placed hi> hut at Hut Point ;
the hest of all sites. since lie coiild here easily reach the st~~itherii Ice Shelf
( PIii)tc~1 ) . 111 1908 Shackleti~n's4iip could not >ail farther miitli than Cape
Downloaded by [New York University] at 06:16 15 February 2015

PIIOTOI.-Tlie 1902 H u t huilt hy Captain Sriitt. ahtrut 14 miles suutli uf the 1910
Hut. Behind is the low peak of lava lino~\nas Ohservatiun Hill (710 feet).

PIIOTO 2.-TIie 1010 H u t built by Captain Scott uii 1laclIurtlri Sound. Sotice the
the right. huilt of food ca3es. Behind is the lava scarp callrd the Ramp. and
. \ i i i i e ~ 011
the volcano o f Erebns ~13.000ft.).
 19421 CAPE EVANS, ANTARCTICA 9

Royds, where he was isolated from the mainland by open water, or by the
crevasses of Erehus during the best sledging months.
O n Scott’s second expedition, in which the writer was senior geologist,
we found thick bay-ice south of Cape Evans on the 4th of January. This
fixed tlie site of the hut a t Cape Evans (Photo 2 ) . Tlie southern parties
were able to reach the Ross Ice Shelf as late a s the 24th of January, but in
a few days open water separated the hut froin the Ice Shelf. I t is of interest
that our sledging parties returning from the south were unable to return to
our o~viihut at Cape Evans until the 13th of April. I n the meantime we
lived in the coinniuiial settlemcnt at H u t Point (Photo 1 ) . Two years later
two men were drowned on this short crossing of 14 miles, because they tried
Downloaded by [New York University] at 06:16 15 February 2015

to traverse the sea ice to Cape Evans before the new ice was thick enough.
In the map (Fig. 2 ) I have added actual dates in 1911 when the sea froze
near H u t Point ; and when the ict. I)roke away near Cape Evans.
Such then was the uiiiland of Cape Evans. T o the east extended uniiiter-
rupted ice slopes to the suinniit of Erehus a t 13,000 feet. I n places fields of
crevasses prevented any crossing of the land ice; and of course the level sea-
ice was far preferable for sledging, except in the season when it was liable
to break away under one (iii the suniiiier) .
The third question to answer is “\l’Iiat kind of a settlement developed on
tlie chosen site?” I have ventured to talk of Cape Evans a s a “coniniunal
hamlet” ; since it reminds one of the similar settlements in Jugoslavia in the
late iiiiddle ages. (Still better would the term apply to tlie more complex
hamlet a t Little America.) Tlie main liut (Fig. 2) had a floor area of 50
feet by 25 feet, and a simple roof with a central ridge. Three small windows,
permaiiently shut and with doubled panes, adniitted suficieiit light in the
summer, while in the long winter an acetylene illant in the porch g-<tve us
light.
Of considerable interest were tlie precautions to keep out the cold. In
additioli to five layers of \vood and packing forming the walls, the liut was
protected by side-sheds. These two annexes are clearly s1ion.n in the plan
(Fig. 2 ) . O n the west tlie stables were built of coinpressed coal while on
the east an enclosed veraiidah was huilt of boxes of food, etc. T o serve the
varied needs of the 25 settlers there were smaller buildings in the vicinity ;
such as the magnetic hut and the ice grottoes for storing food (and for
pendulum work). There was also a great store of compressed chaff for
the support of tlie ponies ; while tlie clog kennels formed a lively faubozwg
to the south.
At the end of the second winter the southeast blizzards had covered
everything, so that the liut was alniost invisible. But life inside tlie hut,
with the temperature held at 45”, might be sumniarised in the phrase “dark
 10 TAYLOR-GE NET1C URBA hT G E O G R A P H Y [Mar.

and dirty, but comfortable.” Outside the average temperature for J ~ i l ywas
around - 20” F.
If n-e use the analogy of living organisms in connection with this study
of town development, then the hamlet just considered may be likened to an
extremely simple animal consisting of only one cell. All functions are carried
on in this cell, and no “zones” are present. I t represents the simplest of
human clusters, and is clearly in the iiifantile stage of town development.
A few further notes on the environment may be added. The normal
sledging season extends from about September 20th to March 20th. I n
these latitudes (774” S.) the sun vanished about April 22nd and returned
on August 22nd. However the great cold and the almost constant blizzards
Downloaded by [New York University] at 06:16 15 February 2015

made March and September very unpleasant for such sledging as had to be
done. S o vegetation (except rare moss atid lichen) is found in the whole
continent, so that no aniiiials can live on the land. Practically all the food
(except seal meat) had to be brought from civilised lands.
The density of huinaii settlement under these conditions is too low to be
charted. I t may be noted that the nearest similar temporary settlement was
the hut 400 miles north at Cape Adare; while Framheini (Little America)
was about the same distance to the east. As far as the Antarctic continent
is concerned, the geographer can say that the environment is supreme. Man
(in tlie form of explorers and scientists) may choose to inhabit certain minute
clots therein temporarily and a t great expense, but tlie writer sees no argu-
ment in favour of Possibilisin in our first example.

2. Bergcti, a large town i9a the cool temperate portiota of Europe

As noted earlier, our examples are to be taken at average intervals of
10 degrees of latitude. Hence our second example is in a cool temperate
climate! but is not typical thereof; since the coast of Norway has tlie highest
winter anomaly in the world. However it is an extremely interesting town,
for a variety of reasons. Owing to its position on the northwest of Europe
it experiences almost constant, warm, moist, southwest winds. These pro-
duce a typical marine climate (Cfb, as shown in Fig. l ) , with a heavy rain-
fall and a very mild winter.
In addition to the unusual climate, Bergen has an unusual topography.
Owing to the epeirogeiiic uplift of the Caledonian Folds in Norway during
late Tertiary times, the hinterland of Bergen is a plateau of about 4000 feet
elevation. The Caledonian folding took tlie form of a series of parallel arcs,
whose anatomy is clearly visible in the topography today, for the uplift con-
sisted of alternating strips of hard granitic material and of softer Silurian
schists. Erosion since the uplift, whether by water or ice, has removed the
softer schists much inore quickly than tlie harder eruptive rocks, aiid these
 19421 BERCEN, NORWAY 11

“Bergen Arcs” (crossed by what seem to be tension cracks) account for the
regular valleys and gulfs which have in large part determined the site of
Bergen (See the inset in Fig. 3 ) . Thus By, Sor, Kross, and Fuse Fiords
Downloaded by [New York University] at 06:16 15 February 2015

FIG.3.-Plan of the town of Bergen (Lat. 60” N.) in Norway, to show the control
of the growth of the city by the topography. Based on H. Slanar. In the inset the
site of Bergen (black) at the centre of the fold-arcs and fiords is charted.

are hollows eroded in the arcs of soft schist ; while Oester and Voss Fiords
are perhaps in part cross-graben due to tension.
Let us now answer our standard questions. “Why did folks settle at
Bergen in the first place?” W e may safely assume that it was but one of
numerous sites used by the early Viking seamen. I n the soft silts at the
north side of the By Fiord, they dragged their boats ashore near the Sandvik
of today (Fig. 3 ) . In the early days Lungegaard Lake was connected to
Vaagen Bay; while two morainic ridges separated Vaagen from Pudde
Fiord. The town was founded near the present Bergenhus (No. 2 in the
plan) about 1075 ; and in later times several forts were built hereabouts to
protect the harbour.
The Danish King Canute controlled Britain, Denmark, and Norway ; and
he placed his capital at Bergen ; which lay midway between Edinborough,
Opslo (Oslo), and Nidaros (Trondheim) . The name Bergen is derived
 12 TAYLOR-GENETIC URBAN GEOGRAPHY [Mar.

from Berg-vin, i.e., niouiitain meadows, and refers to tlie abundant grass on
the slopes of the By Fiord. During the 12th century English traders were
gradually replaced by Germans belonging to tlie Hanseatic League. During
tlie next few centuries they built large xirehouses all along the north side
of the Vaagen, which is called Tydske-brygge (i.e., Deutscli Quay) to this
day. About 1600 tlie polver of tlie Haiisa declined, and trade had all passed
into Norwegian hands by 1750.
Before examining further the growth of Bergen it may be well to point
out that it is tlie most northern city containing 100,000 people. This is the
iiiore remarkable since it is not a capital city, having been replaced by Oslo
several centuries ago. (Helsinki. Stockholm, and Leiiingrad are all further
Downloaded by [New York University] at 06:16 15 February 2015

south. j S o r has it any fertile hiiiterlantlt as is obvious from the inset map.
\Ye may safely say that its advantageous position in respect to tlie fishing
grounds, and to tlie trade of the Sortli Sea and the Atlantic Ocean, led to its
very considerable expansion. As the map shows, it is protected from storms
11y the “skerry guard” of the low Sartor Isles. Areas of farimland on the
floors of the nunierous glacier-cut troughs are relatively large to tlie soutli-
east of Uergen. Especially after tlie Bergeii-Oslo Railway was built in 1900
did the city grow vigorously. Fish at first, and later shipbuilding and an
entrepot trade in grain, wool, and petrol, account for Bergen’s prosperity.
It is still the chief trading port for Norway.
Reference to the main niap in Figure 3 shows that Bergen was confined
to the east side of Vaagen until 1300. This was the Hanseatic centre, per-
haps a t its maximum about 1500. Then the swampy land at the head oE
Vaagen was gradually covered with streets and houses. The present shop-
ping and office centre is along the Torv (market) and Strand streets. Here
also is one of tlie largest fish markets in tlie world. By 1750 the area
between I7aagen and 1,ungegaard was all occupied by buildings (Photo 3 ) .
Bergen has been tlie scene of many fires, one of the worst occurring in
thc district around tlie Pool in 1916, but this area has all been covered with
large stone buildings since that date. Modern suburbs have spread round
Pudde Fiord ant1 Lungegaard. Various factories, mostly connected with
ships or fish, have sprung up in outlying areas like Laksevaag or Sandvik.
The steep sides of the hills. Floi-Fell (900 feet) on tlie east, aid Lovstaken
(1500 feet) on the southwest, hare prevented the city from spreading far
from the original site around lraagen.
l y e can clearly see the part played by topography and climate in tlie
growth of a town as large as Bergen in this high latitude ; the same as that
of South Greenland, north Labrador and tlie Bering Sea. Yet many cul-
tural factors are of course present. Bergen was an early centre of religious
and political power, and these are man-made factors. Water-power, fish-
eries, tourist trade, and railways depend partly on physical, partly on cultural
 1942] BERGEN. NOR\VAY 13
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO 3.-Tlie Wharf at Bergen. once occupied by the Hansa iiierchatits. The
harbour (Vaageii) lies to the left. (Courtesy of Robert S. Platt.)

conditions. .4 determinist nil1 stress the former and say that water-power
depends primarily on topography antl rainfall, fisheries depend n i l the teiii-
perature and food conditions in the sea. Tourism flourishes where Nature
has displayed her choicest scenery. Kailways inust go where the natural
corridors are easiest. Man can choose one of an infinite iiumber of ways oi
exploiting a country ; but it seeiiis ti i the writer that our “cIiooser’* rather
resemljles the so-called “yes-man.” If he is to prosper, lie does well tii
chr~osewhat Dame Nature the Dictator indicates is the best for him ! !
nIy visit to Bergen in 1931 was too short for me to make a functional
survey of the town : and so far as I !inow this is not availahle. Many of the
historic data in my main niap are Ixtsed on H. Slanar (PYOC.Grog. Soc.,
Vienna, 1918). It is obvious that the city exhibits the usual plaii ; with the
lmsiiiess centre near the Torv, antl with niore or less concentric zones of
offices, apartments. antl villas t o the south. These zones are somewhat Iiorse-
shoe shaped (rather than circular), in accord with the unusually hilly
topography of the site.

3. McAfiirray, a fur-trodirrg river port iri North driicrica

In 1778 Peter Pond explored the region north of Lake Athabasca, and
about 1790 the Northwest Company established a fur-trading station at
McMurray 011 the Athahasca River (Fig. 4 at A ) . It was situated on tlie
iiiain corridor of canoe-travel from the Great Lakes to the Mackenzie River.
Canoes reached Lake \Vinnipeg rv’a tlie Lake of tlie Woods. Thence they
 14 TAYLOR-GENETIC U R B A N GEOGRAPHY [Mar.
Downloaded by [New York University] at 06:16 15 February 2015

FIG.4.-Settlement conditions in the Taiga of north Canada. At A is a map of

the Mackenzie Basin, showing NcMurray and Waterways as well as the La Loche
Portage. At B is a block diagram of the twin towns of McMurray and Waterways.
-It C is a detailed plan of the “infantile” settlement of Waterways. At D the same for
McMurray.

paddled north to Cumberlaid House on the Saskatchewan, where they

entered the Sturgeon River. At the head of this stream the Frog Portage
led to the Churchill River which rises in La Loche lake. At the north end of
 19421 MCMURRAY, ALBERTA 15

the latter was the famous La Loclie Portage where Netliye Fort was built
(Fig. 4 ) . A rough wagon road of 124 miles was constructed here in 1875,
so that heavy goods could be placed on tlie Clearwater River.
McMurray, the settlement under discussion, grew up 40 i d e s west of
the portage at the junction of the Clearwater with the mighty Athabasca
River. The waters of the latter, after passing through Lake Athabasca and
Great Slave Lake, reach the Mackenzie River. The first steamer traversed
the Mackenzie from Fort Smith Rapids in 1887, while in 1920 tlie railway
from Edmonton reached Waterways on tlie Clearwater.
Today the train traverses the Taiga north of Edmonton once a week,
and I made niy survey in July, 1936. At Lac La Uiche (132 miles north)
Downloaded by [New York University] at 06:16 15 February 2015

is the last of the farming country, though small patches of crops are grown
right to the Arctic Circle. The chief inhabitants north of Lac La Biclie are
Cree Indians, engaged in trapping and fishing. The little towns, marked
hereabouts on the map, usually contain fewer than half a dozen houses. I t is
therefore somewhat of a surprise to find that there are two towns at the end
of the railway. These are Waterways and McMurray (Fig. 4 at B ) .
Possibly it was intended to carry the railway beyond Waterways to tlie older
settlement of McMurray. But deep water is available in tlie Clearwater;
while rather high prices were demanded for the necessary lands at Mc-
Murray, so that there seems no likelihood of the railway being extended at
present over the intervening three miles. Waterways controls the railway
traffic, and here the loading of the steamers and scows for the northern
country takes place. McMurray is still the centre for the local fur-trade,
and for the growing air-borne traffic to the mining fields of the north.
There is little doubt in the writer’s mind that this “twin settlement” is
destined to grow rapidly in the next few decades. Hence it seems of impor-
tance to learn just how the town has progressed in the years since the railway
arrived. One must confess that Waterways shows no evidence of any
design whatever (Fig. 4 at C) . The railway curves round the station iii a
loop which serves the huge storage sheds of the Hudson Bay and Northern
Transport Companies. The offices of tlie Companies are naturally close to
these sheds. A large oil depot is between the two (Photo 6 ) . Five small
steamer-scow units link Waterways to the portage at Fort Smith. The river
is open from mid-May to mid-October.
There are a dozen wooden houses, mostly little more than frame shacks,
scattered in the vicinity. A few hundred yards froin the station is the com-
bined hotel and store-a more elaborate two-storey buildiiig-which is tlie
iiatural centre for Waterways. Water for the hotel is obtained from the
village pump across the street. One other smaller store is nearby. There
was (in 1936) no church or school in Waterways; indeed about a score of
wooden bungalows and shacks were all the other buildings in the place.
 Perhaps one r e a m i for the lack of plan is the possiliility of floods. I n
the spring antl autuiiiii th- riorthern I)arts of the rivers a r e frozen, \vliile
the s c d i e r i i waters are still pouring their supply northwards. Hence wide-
spread f l o i ids occur at times, which tlrowii the houses of \\'ater\rays in
several feet of \rater. Back o f the settlement (Fig. -I) are rounded hills
rising 300 feet al)ove the river. T1ie.e are clothed i n a regular jungle of
conifers antl aspeiis. Xatural clearings with consideral)le areas of pasture
occiir in places, aiitl m e such grassy meatlo\\- was grazetl l)y the cattle of the
earl!. lirr traders (about one mile east o f Mchlurrayj i i i the district iiow
called Prairie (Fig. 4 at B ).
Fcir a settlenient 140 years c i l d MclI~iri-ayis not very imposing. There
Downloaded by [New York University] at 06:16 15 February 2015

is the straight main street al,out a mile l(ung, \vhich runs \vest to the junction
of the two rivers (Fig. 4 at L) ) . All tlie hriuses are slionn in this large-scale
map ; and they are aliiiost ;111 fraiiie h i u s e s antl one storey high, tliough the
slirq's often have a lake two-storey front. Hoivever tlie Hutlsc.m Bay Store
antl the chic+ hotel are iiiure imposing I)uiltlings. atid most i ) f them were neat
antl freshly I)aintetl. Indeed the schc i(11 antl chtirches are quite attractive
buildings (Photo 4).

PHCITLI -I.-The main street of llr?rlurrdy ( ;\herta) luokiiig to the west. The
large huilding on the left is the Hotel. the Hudson Eay Store is four buildings to the
west.

In atlditicin to the Iiig C'~J1ilpali~

store there were three other stores, three
cafPs. two I)utchers, tux) Imrbers, ant1 a drug store. The Batik, Post Office,
and Dance Hall are in the shopping Mock. At the west end near the l i g
river is the \\.ireless Station, \rliile a few fur-tlealers indicated the main local
product. In the detailed niap an area o f several acres is showx as grcnving
potatoes. These I,elongetl to a Japanese farmer, and seeiiietl to be flourishing.
 19421 h I C M UKRAY. ALBERTA 17

The air-transport offices are on a hranch of the Clearwater called the

Snye (Fig. 4 at B ) . I was iuuch interested to find that their haphazard
arrangement along the bank of the river was due to each office being mounted
on runners like a sled, so that during the flood periods in spring and autumii,
they could be drawii t o higher land to the south (Photo 5). This is a form

ramma+* &iL*
i
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO 5.-The ( J ~ ~ ~ CofC Sthe .lir Translwrt uii the Clearwater River near I I c l I u r -
ray. These offices are niouiited uii runners, so that they can be dragged to higher
ground in the spring and fall floods.

of traiisliuinance new in m y experience. tl\t the time of my visit the three

air coiiilmiies \\‘ere shuttling ten planes to and froin the nnrthern iuining
fields. ncitalily tlitrse near Yelhnvkiiife. Finally, al)out a mile from Mc-
Murray i i i the high shale cliffs crf the .AtIial)nsca are to be famicl thick layers
uf Tar-sands, which in the future will be used as sources of petroleum.
RIcRIurray is I Iwlieve the sole tcnvii north of the Xll~erta-Saskatclieu.an
fariii lancls. But Cliurchill and M inie of the Yukon towis are doubtless
larger. IHuwbeit it has few rivals as a high-latitude town in Korth America.
It was fnuiitled aliout the same year as Toronto, I)ut while the latter has
growl1 to a city id XOO.000, hIcNurray has a popdation of aliout 600 in
simmer and 300 in ninter. Its satellite 12’aterways supports allout 300 folk
in suiiiiiier, but the Impdatioii falls ti, al)out 30 in winter. Man could
“choose” to go there in his thousaiitls, h i t for environmental reasons he pre-
fers Toronto and XIontreal !
McRIurray hoxever is large enough t o show the beginnings of func-
tional zones. The oldest houses are at the river end of the main street, but
the business centre is the Iilock between the hotel and the HII~SCJII Bay Store.
 18 TAYLOR-GENETIC U R B A N GEOGRAPHY [Mar.

The churches and schools have been built in clearings cut (nit of the birch
forest to the east. There is not yet any tlifferentiation intn first and second
class residence zones. so that it may be classed in n.liat later is called the
juveiiile stage (1)age 49). 12'aterways is too iiiimature t o sliinv any plan,
aiitl is in the infantile stage (, Pliotri 6 ) .
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO 6.-The oil tanks at Waterways. The frame building is typical of this
pioneer settlement. and was transferred here from a former rail terniinus.

The site of McMurray is clearly tletrrminetl 11y the junction of the two
streams of cancne traffic via the .\tlialxisca and La Loche. The steamer port
\vas placed here because of the ral'ids in the ;\thabasca just above Mch2urray.
The nest siniilar steamer port is at Fort Smith Rapids, almit 250 miles to
the north. But the railway is unlikely to be extended t o Fort Siiiith since
air traffic is already reducing the importance of steamers and railways in the
empty northlands. ,4part from its juncti[in-ptrsition I)elii\v the rapids, Rlc-
Rlurray seems to offer no advantage over a dozen other sites iii northern
central Canada. But this definite eiivit-oniiiental factor has given it a small
lead, enough to make it the railway terminus and air-port. Cannot we agree
again that iiian has "chosen," out of many "possi1)le" sites, that which Sature
had already niarketl out as the best ?

11, ( I ddttr city itl tht- fc-rrrpc'rt7fcZ011t' Of .-lsitr

Pekin offers a fair example o f a large city which has developed under
teiiiperate comlitiniis iii a gigantic deltaic plain. It is larger, older, and much
more complex in its evolution than any of the other examples, so that it
represents the other extreme in our stiitly froiii the settlement at Cape Evans.
 19421 PEKIN, CHINA 19

Yet we must try to answer our fuiidamental questions just the same. I t is
difficult to point to any enviroiiiiiental factor which has led to Pekin’s domi-
nance over most of the towns in the deltaic deposits of the Hoang-Ho. These
cover an area of about 80,000 square miles; for though the plain exteiids
south to Shanghai the southern portion (containing this city and Nanking)
does not derive from the Hoang-Ho.
One would have expected the chief northern city to have developed either
iii the centre of the plain, or near the main river, or possibly at a good
harbour on the coast. Pekin fulfils none of these conditions. In early
historic days the capital of China was undoubtedly in Shensi. About the
time of Christ it was moved east to Honan, and about A.D. 317 to the coastal
province of Kiang-su. For a time around 960 it was placed a t Kaifeng,
Downloaded by [New York University] at 06:16 15 February 2015

which has a central position on the great river. However in the 12th century
the Kin tribes of the north made Chung-tu their capital; and this town
occupies the site later known as the Chinese section of Pekin (Fig. 5). The

FIG.5.-Various maps showing the evolution of the city of Pekin in the plains of
north China. A t the left is shown the evolution of the city (from the Ertcyclopedia
Sinica) . A small map is inserted ,to show the deltaic plain, and the roads and railways
to Pekin. A generalized map of the “concentric cities” of Pekin is given on the right.

Mongol conquests resulted in Kublai Khan, about 1280, building a new

capital (Khan-baliq) iimnlediately to the north of the older Kin city of
Chung-tu. This was the capital described by Marco Polo. Towards the
end of the 14th century Nanking was made the capital, but from 1403 to 1911
Pekin continued to be the capital city of China.
 20 TAYLOR-GEXETIC U R B A N GEOGRAPHY [Mar.

It seeiiis clear that a considerable "human" element is iiivolved in tlie

choice of the site of Pekin. In early clays necroniancers ascribed to the site
a peculiarly fortunate character. No douht tlie chief factor was political.
since here the Chinese ruler could keep an eye on the marauditig Mongols ;
while later tlie iJlongcrl coiiq~ierorspreferred to rule the Chinese froni a city
not too far away from their o\vn nati1.e land.
The tra\.eller reaches Pekiii across a vast sandy plain, and 110 river or
canal links Pekin t o the miter world. The p a t , hefirre the railway era, was
a little to\w o i i the Pei Ho River to the east. Today Tientsiii is the chief
entrepot for tlie capital, aiitl the writer reaclietl it hy this route in November.
1926. From a distance, the walls o f Pekin give it the appearance of a
Downloaded by [New York University] at 06:16 15 February 2015

gigantic 111IX,with sides aliout live niiles sqmre-\\hich has been flung down
ul~oiithe plain. Thc presetit poptilatic 111 is nlioiit 1,600.000, but about Iralf
tlie Chinese City of toclaj* is only sparsc'ly rrccipit.tl hy ho~ises,the southerii
portion coiitaining iiiany teiiiple gardens antl large cq)en spaces. Ancient
moats wrrc )mid I N )th portions of the city, \ M e frc )iii the walls the iiuiiierous
trees in the palace gardens am1 tetiil)les inakc ;I picturesclue nntl unusual scene
in this lnre antl level part (of north China ( Phuto 7 ) .

P H O T O 7.-.\ view o f the Great \\'all and tilie of the typical gates 011 the sotitherti
side of the Tatar City of Pekiii (Peipiiig).

Turning t o actual plan of this Oriental city, today its arraugcmt.tit is

unique. The Chinese city consists ()f the usual tiarru\v crowded streets CJf
an Oriental settlement, flanked by innuineralie small shops which cater
largely to the tourist tratlc (as some of the street iiaiiies suggest). The walls
 19421 PEKIN, CHINA 21

round this quarter were built in 1453, and are not quite so high as those
around the Tatar city, which are 50 feet high and 40 feet thick. I n the latter
the Etreets are broader and the buildings larger and much more imposing.
The Foreign Legations occupy the southeast portion of tlie Tatar city be-
tween the Hata Men and Chien Men Gates. They are enclosed in high
private walls which enabled them to withstand the Boxer rebels in 1900.
A stream formerly crossed the Tatar city from North to South; and this
has given rise to a number of ornamental lakes, which beautify the palace
grounds of the Imperial City. Tlie latter is enclosed in tlie centre of the
Tatar city, and in turn it protects the Forbidden City where the Eiiiperors
spent most of their lives. In the northerii part of the Iiiiperial City is a
Downloaded by [New York University] at 06:16 15 February 2015

sniall hill 200 feet high called Coal Hill. It is crowned by several small
temples, and contrasts with the surrounding plains ; but this “eminence” is
undoubtedly artificial in origin.
The Tatar City is covered with buildings, streets and palaces ; and though
many Manchus still live here, the majority of the people are now Chinese.
I t is difficult to chart functional zones for this city, for the ancient and modern
plans are somewhat at variance. Originally the emperor’s palaces and house-
hold occupied the Forbidden City. Tlie government officials lived in the
Imperial City, the ruling caste oi Maiichus in the Tatar City, while tlie sub-
ject groups of Chinese were in the adjacent but separate Chinese City. To-
day the palaces are largelv iiiuseunis, while the government offices are iiiostly
scattered through the southern half of the Tatar City, though soiiie are still
in the Imperial City. Probably tlie huge walls will be levelled in the near
iuture, and the space converted into wide boulevards and gardens. It may
be noted that though large areas oi the Chinese City are empty, yet there are
a few small “suburbs” outside the walls, especially to the east of the Tatar
City.
A glance at the map of China shows that there is no other large city in
North China near to Pekin, except Tientsin. The iiiagical and political fac-
tors (briefly referred to earlier) led to the city’s birth, perhaps as far back
as 723 B.C. A t this time “Chi” was the capital of the Yen kingdom, accord-
ing to the Encyclopedia Siizica (Fig. 5 ) . Given this start, no other city
seems to have arisen to compete with it, though other sites such as Kaifeiig
would seem to have more to recoinmend them. I n process of time the main
roads inevitably led to Pekin, and six of these Imperial Roads ( A to F ) are
shown in Figure 5. Later the four main railways also radiated from Peltin,
respectively to Mukden, Paotu, Canton and Shanghai.
I n latitude atid origin Pekin reminds one of Madrid, but the climate is
rather different. There is a hot, wet suninier and a very cold, dry winter.
 22 TAYLOR-GENETIC U R B A N GEOGRAPHY [Mar.

Owing to its eastern position in the huge laiid-mass of Asia Peltiii lias a
rather extreme climate, as is clearly shown by the hythergrapli in Figure 1.
The long extension of tlie graph to the right (iii suiiiiiier) indicates a marked
rainy season, so that the graph for Pekin is fairly typical of Monsoon coii-
ditions.
The present plan of Pekiii is essentially orieiital aiid medieval. But an
enclave of occidental culture is being incorporated in it, especially in tlie
southern part of the Tatar City. W e may perhaps compare this enclave to
the iieotechnic sections developing in certain parts of an occidental city. (1
return to this concept later when discussing tlie plaiis aiid growth of Trento.)
The palaces clearly form the nucleus of Pekin, even though today no emperor
Downloaded by [New York University] at 06:16 15 February 2015

lives there. Arouiid them is an adiiiinistrative zone, and beyond that the
upper-class town of the Tatar City. Smaller shops aiid slum areas are rather
characteristic of the separate southern quarter called the Chinese City. 1
shall not try later on to give a forinula for Pekin ; but may remark that we
are here confronted with an example of city structure as complex as some of
those coiiiposite miiicmls, whose chemistry aliiiost defies tlie forniula-maker.
Luckily few towns are .so complex.
It inust be adiiiitted that the choice of the site of Pekin is not due to any
iiiarked environniental factors. Obviously in such a rich agricultural region
as the Hoaiig-Ho delta, soiiie large distributing centres were bound to de-
velop, but they are less characteristic of a primitive oriental culture than of
a modern industrial occidental culture. We may perhaps grant that Pekin-
situated in a vast region of iriiifoviiz environment-is a good exaiiiple to sug-
gest that the possibilist theory does, under such conditions, explain the facts
of geographical distribution.

5. Bisfrn, a desevt towvz iiz'tlzc Meditcl-rmteait regioiz

In 1938 I was able to carry out a life-long plan to coinpare tlie Sahara
Desert with the arid areas of Australia. While the environiiieiits are alike,
there is naturally very little resemblance in the human settlements. This
may, a t first sight, be taken as a triuinph for tlie Possibilist Party. I aiii
of the opinion, however, that the dissiinilarities in the two types of settle-
ment are iiierely due to the different stages of development obtaiiiing in
Algeria and Australia.
Let us compare Coward Springs in Australia (450 miles northwest of
Adelaide) with Biskra. Both lie on the edge of the desert, and have a
winter rainfall of 4 inches. The coveritlg of sparse but regular vegetation
in each case is similar in pattern though iiot in genera. Both, curiously
enough, are situated on the edges of artesian basins. Both have railways,
which have been running for inaiiy years, in the case of Coward Springs
 19421 BISKRA, ALGERIA 23

since 1890. When I visited Coward Springs in 1919 it was tlie chief settle-
ment on the railway for SO miles in either direction. A sinall date oasis had
been planted here many years before, yet the ‘‘town” contained only four
houses, two of which were empty. (Given an excellent environment, such
as at Chicago, a settlement might add a million to its population in the
same period.)
Let us now see what has happened in the Sahara with a far lower
standard of living and a far greater population pressure. I n spite of an
equally unfavourable environment, Biskra is an apparently flourishing town
with about 10,000 inhabitants, of whom 2000 are French (Fig. 6). I t is
Downloaded by [New York University] at 06:16 15 February 2015

2 0”

lon
4“

P O PU LATlO N

FIG.6 (I&).-Site-view of Biskra looking to the northeast. The town extends

half a mile southeast of the station. The church, chief hotels, post office and town
hall are indicated by initials. The chief Arab quarters are black. F. is a factory site.

FIG.6 ( r i y h f) .--Sketch map showing population density per square kilometre.

Heavy lines are isohyets. (Both” maps from Grog. Rev., April, 1939; isohyets added.)

supported primarily by the date oases, but also by the Freiich niilitary sta-
tion, and to a lesser degree it acts as a caravan teriiiinus and (in winter)
as a tourist resort. My own visit was made in midsummer; when, it can
readily be understood, our appearance was hailed as manna iii the wilderness
by the hordes of touts aiid beggars.
Biskra lies 25 miles south of tlie striking gateway through the Aures
Mountains called El Kantara. Even in Konian times there were garrisons
at El Kantara and to the south at Veskara (ic., Biskra). Hence the town
has had a long and eventful history.
 24 TAYLOR-tiENETIC V RHA N G E O G R A P H Y [Mar.

The little river Tilatcou runs south from the interior plateau to the
marginal range, and thence through the gorge at El Kantara to the Sahara.
At this settlement wheat is grown cni little hill-fans just north of the gorge ;
while south of it in ur near the stream-bed is a large date oasis \vhich sup-
ports 6000 -4rabs. From El Kantara the water (in winter) flows south
for iiiilea into the veritable desert atid reaches Biskra (Photo 8). Thence
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO &-The palm Oasis of Biskra in the Sahara, showing the water from the
Atlas hfountaitis which irrigates the Date Palms.

it flows south, and, at times, reaches the salt playa called Shott Melrir. nhich
is below sea level. Small mesas diversify the desert just north of Biskra.
Oleanders grow along the dry hetl nf the stream with occasioiial patches o f
date palms. The vegetation is in tussocks with usually low tanlarisk clumps
or masses of reeds.’
7 “Sea to Sahara,” Griffitli Taylor. Gro!/. Rw.. 3 (April. 19.39)
 19421 BISICRA. ALGERIA 2s

There is little of the pre-French town remaining at Biskra. =\ rectangular

arrangeinent of streets, with the iiiaiti axis running N.W.-S.E., extends for
half a mile between tlie station and the group of large hotels in the south-
east. The centre of the town contains the Town Hall and Post Office, while
barracks a r e to be found near the wadi to the east. But the most note-
worthy feature of tlie town is the large Fort (400 Iiy 200 yards) which
defends the town on the north. This is surrounded by stont Iirick walls
about 20 feet high, and guarded by tower5 with slots for rifles. Inside this
wall a r e the riijlitary offices, Iiospitals, as well as many siiiall houses for
officers. etc. (Photo 9). *At the extreme southeast corner of BisLra is

i i
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO ().--The walls of Riskra. with a flock of goats brought in from the environs.
.A tower of the French Fort appears in the background.

the small “negro village,” though the dwellers seemed to he of =\ral, rather
than iiegro origin. A large and roughly circular area of date plantations
extends for nearly 2 miles to the south of the town.
T h e houses are built of brick or of irregular stones covered with
stucco and painted white. In tlie town Iwth tlie French atid Arab quarters
contain two-storey houses, but in the lower-class houses in the quarter south
of tlie main town, the characteristic high Arab walls enclosed small yards
011 to which various flat-roofed rooiiis opened. .A crowded open market,
as well as a mosque. a r e typical I>uildiiigsin the Arab portion of the town.
Biskra is the terminus of a light railway which ruiis south to Tntiggoiirt,
and collects tlie date harvest for export from the Sahara (Fig. 6 ) . Possi-
 26 TAY LOR--GEN ETIC L’RBA N GEOGRAPHY [Mar.

bly foreigners imagine that iiiauy such railways may run south into the
rainless desert iii the future. But it is tlie artesian water which makes
possible the string of oases to the south of Biskra; and so far as I know
there is 110 other locality in tlie French Sahara so well endowed. I n pre-
historic times tlie rains on the Ahaggar Plateau gave rise to the large river
Igharghar, which flowed north to Touggourt. Probably today some of tlie
scant\- waters flow far to the north along relatively deep-seated periiieable
beds, and repleuis11 tlie artesian basin uutler Touggourt. Yo doubt waters
from the ‘Atlas mountains also reach this basin.
The city-plan of Biskra shows us a fort in tlie north with a park belt and
a zone of offices just to the south. The latter lies in the centre of a belt of
Downloaded by [New York University] at 06:16 15 February 2015

better-class French houses. Still farther south is another zone of second-

class houses, while yet south again is a belt of smaller oriental Arab houses.
How does Biskra fit into our determinist-possibilist debate? When we
think of the enornious length of tlie northern border of tlie desert, which
extends ahout 2500 miles from the IVacIi Draa to Suez in the east, it is clear
that onl>- a few districts are as favoured a s Biskra. There is only one by-
gone Ighargliar River, and probably no equal to the Touggourt artesian area.
Figuig and Laghouat are similar desert outposts, but only the formet is
served by a desert railway. Hence I should make use of our Biskra example
as follow :
Given desert conditions such as obtain in the Sahara, there are only
a very few sites where the ameliorating factors have justified modern enter-
prise in developing the district as the French have developed Biskra. Even
the military site mas strictly determined by the water and the traffic con&-
tioiis. \Ye may note that these were also in operation in Roiiiati times,
centuries before the French took over Algeria. As usual I see Nature (i.e.,
the en~ironiiient)acting as tlie dictator far more than man in the settlement
problem under these difficult conditions.
The 11ossibilist, it seeiiis to me, puts tlie cart before tlie horse. H e would
say “Ah, but the vital trade in dates owes much to the artesian bores and to
the desert railway; and surely these are due to human energy?” The
deterinjnjst replies “Man can put down bores and build railways anywhere
in the Sahara ; but in the vast majority of cases he takes very good care to do
so oiily where Nature has provided the conditions to make such expenditure
worth while.”
In such exploitation Nature determines the route of development, inan
only deteriiiines its rate. T o return to the parallel with Coward Springs in
Australia. I see no sign of mati utilising this southern region yet. But
when population pressure has increased to soiiiething like that present in
Algeria, then we shall find man i n Australia using the methods of develop-
 19421 URANDAN Jl, QUEENSLAND 27

meiit which Nature has driven inan to use in tlie same environment at
Biskra.
\lie may illustrate this phenomenon of population-pressure by compar-
ing the population densities in three regions, each of which has much the
same environment. They are western Spain, Algeria and southern Australia.
All have a Mediterranean climate with siinilar and pleasant temperature
conditions on the whole. I n all three regions the rainfall diminishes in the
interior. I t is obvious that crop-densities (and therefore populations)
depend primarily on rainfall.
In the graph inset in Figure 1 the relation is expressed approximately by
a straight line for each of these regions. Spain and Algeria exhibit very
Downloaded by [New York University] at 06:16 15 February 2015

similar conditions ; and, to the writer, this indicates that tlie population-
pressure is about tlie same in each. Southern Australia shows a much lower
density of population for tlie same rainfall. A t a rough approximation, this
seenis to indicate that the population pressure in Spain and Algeria is about
eight times as great as in southern Australia. The natural result is that
standards of living are far higher in Australia. But it is entirely probable
that population pressure in Australia will increase considerably as time goes
on. Later on, the graph-line for Australia will climb to a position much
closer to those for Spain and Algeria. The difference is not really due to
human choice, but to the i ~ ~ z / i m t z wcharacter
e of Australian settlement.

6. Urandaizji, a senzi-desert Izaiizlet in tlzc east of tropical Azistrnlicr

For iny sixth sample I have chosen a tiny village on the pioneer fringe
of Australia, of which I made a sketch survey when I was investigating the
rival trans-continental railways in 1922. Urandanji, tiny as it is, appears
on most maps of Australia for it is tlie sole settlement in that part of Queeiis-
land. Caniooweal is 120 niiles north, while Boulia is the same distance
to the south. Somewhat nearer are tlie mines at Dajarra and Duchess oii
the railway to Townsville. The striking feature about Uraiidanji's position
however is that there is no town to tlie westward for 1200 miles, until one
reaches Marble Bar not far from the Iiidian Ocean. It is true that there
are a few large cattle ranches near Barrow Creek (Fig. 7) ; but hardly a
score of white folk inhabit these areas of sparse pastoral occupation. They
are indicated as white patches in tlie uninhabited portions of tropical Aus-
tralia, which are shown by close ruling.
What sort of a settlement has grown up in this pioneer region of Aus-
tralia during the fifty years or so of its pastoral development ? With a total
rainfall of only 11.4 inches, and an average temperature of 74" F., one could
not expect much of a human agglomeration, especially as the evaporation
is not far from 120 inches a year.
 28 TAYLOR-GENETIC URBAh- GEOGRAPHY [Mar.

One day in August 1922 we motored west across the rather rough land-
scape between the railway a t Duchess and Urandaiiji. W e passed two
ranches in which the usual bare iron buildings (raised on wooden piles to
combat white ants) rather repel the traveller who conies froin iiiore favoured
districts. Some 20 miles from Duchess we climbed up a stony pass atiiid
giant ant-nests a i d cluinps of desert spiiiiiex grass (TF-iodia). West of this
range the rock outcrops changed froin sandstone to shale, with an iiiiiiiediate
improreinent in tlie grazing capacity. I n these plains tlie soil is often
a red gravelly clay and carries sparse Rlitchell Grass. Wild bustards and
large kangaroos were occasionally seen.
The little township lies 150 miles on the hot side of the Tropic
Downloaded by [New York University] at 06:16 15 February 2015

oi Capricorn. I t consisted (in 1922) of an hotel, two stores, the post

office atid a few other houses disposed on each side of tlie broad main
street. (Fig. 7). 111 all cases the walls are made of “weatherboard” i.e.,
planks cut froin hard eucalypts (\zit11 a wedge section) atid nailed in a

URANDANJI 1922

\
Qt
Sheds

is
____----
T r a c k jii 7ruck 6
Camo a weal 0640
---__ --
-a- -
ce

FIG.7 ( f o p ).-A map of tropical Australia to show the position of Urandanji. just
to the east of the widest expanse of empty land in Australia. Ruled areas are desert,
and isotherms and isohyets are inserted.

FIG.7 (bottom).-A sketch survey of the little settlemellt of Urandaiiji on the

main stock-route from Camooweal to llarree.
 19421 URANDANJI, QUEENSLAND 29

horizontal position. Iron roofs and verandahs are universal. The most
prominent features were half a dozen iron wintlmills, which pump up
river soakage from wells sonie SO feet deep.
Though 110 large trees grow in the township. yet there a r e some fine
hedges of acacia; antl two o r three of the gardens contained beautiful rosr-
trees. The Georgina River rises north of Camooweal, and after a course
of 200 miles passes close to Urandaiiji. In times of flood the river inay be
several miles wide, Ilut on my visit hardly any rain had fallen for the last
six inonths (Photo 10). However there was a fine pool in the river,
about half a mile long, near the township, which indeed owed its position
rartly to this vital factor. the river.
Downloaded by [New York University] at 06:16 15 February 2015

PbroTo IO.-The (korpiiia Kiver \vliich fciws through LTraiidanji i n the \vest c i i
Quernsland. The rainfall i s about 10 i n c l w here, but large waterholes persist for matiy
months iii suiiie parts crf the river.

Further up tlie river in Northern Tcrritciry i b Lake S:tsli (Fig. 7 ) .

It is the only permatietit water availalde to the wild blacks in a dr! season.
antl accordingly they tend to congregate here each Kowiiiber. X largc cnt-
tle station has I)eeii estal)Iisherl at Lake Nash. to tlie west of which is thc
Arunta Desert.
ll’e may sum up by stating that under the enviroiiinerital coiiditionh
cited, with cattle ranches about 20 niiles apart, a little town such as
Urandanji has tlevelopetl a h but hall way hetween Caiiiooweal and Boulia,
i.r.. 120 miles froin either. The hanilet is 011 the great stock-route oi
cattle travelling from Northern Territory and Caiiiooweal to the South
,Australian Railway at llarree. Some (lay Urantlatiji may lie on or
 30 TAYLOR-GENETIC URBAN GEOGRAPHY [Mar.

close to tlie logical transcontinental railway, wliich mill link Darwin to the
populous country in the southeast of the continent.
It is worth noting that tlie Ui-andatiji district is a homocliiiie oi much
of tlie Darfur, Chad, and A’liddle Xiger regions iii Africa. These receive
about 11 inches oi suiiiiiier rain, and this is sufficient to enable liuiidreds of
small towis to develop in Africa. The average populatioii detisity in tlie
-African belt is about 6 per square mile. In Australia it is about one-eighth
of a person per square mile for similar eiivironments. The writer does not
agree that this illustrates “human choice,” i.?., that SO folk choose to live
in Darfur for oiie \vho chooses to live in \Yestern Queeiisland. It is merely
another example of tlie stages of development in iiiaii’s occupation oi a
Downloaded by [New York University] at 06:16 15 February 2015

specific environment. In a pioneer country, with no population pressure

ant1 1%-ithhigh standards oi living, we find an early stage with a density of
onl\- one-eighth.
Uiider the competitive conditions obtaining in the Sudan, six folk per
square mile can be supported at a very low standard of living. Though it
is highly unlikely that Urandanji will ever reach this high deiisity ; its
poIiulation may advance considerably as pressure increases in tlie rest of
A4Listralia. However, a focal resident (Mr. V. Bye) writes iiie in 1931 that
only two iiiore houses have beeii built in Uratidanji in the last twenty years,
while oiie original house has been burnt do\vti and one store closed. Hence
iiiaii has not exercised his powers of choice very noticeably as regards this
possible site.

7. Saiita M a r t a - a tiyopica1 ~~zoiisooiz

town i i i Soirtlz Aiitericn
For niy last exaiiiple I have chosen a sea-level town only 11 degrees froni
the Equator. Here with an average iiioiithly temperature of 81.5”F. we find
a fairly typical tropical settlenient, even though the raiiiiall oi 43 inches
is rather small for such a low latitude. Santa Marta in Colombia is tlie
most northern large town in South America. Moreover it lies south-
west of the first settleiiieiit made in 1493 by Coluiiibus in Saiito Domiiigo.
Since the constant trade-winds would drive the early Spanish ships south-
ward to Santa Marta, it is not surprising that Ojeda built a fortress here
in 1502. possibly before the foundation of Nombre de Dios in Panaiii5.
T h i s our seventh exaiiiple is in a seii,se tlie oldest settlement in North or
South Ahierica. The town however was not properly established until
1525. M y survey was made iii December 1930, when I made several short
journeys into the Andes, wliicli here reach their most iiortherii position.8
;Z rocky cape projects like a hook to tlie southwest, and so affords ample

8 “Settlement Zones in the Sierra Sevada de Santa Marta, Colombia,” Griffith

Taylor, Geog. Rev. 21 (Oct., 1931).

 19421 S A N T A MARTA, COLOMBIA 31

C A R I € A N S E A

Cab0 de la ARuja
Downloaded by [New York University] at 06:16 15 February 2015

ZONES OF SETTLEMENT .
Pai-am0 (glaciated inp/aces)

Old coffee plantations

Jungle and coffee belt
Jungle [some sugar cane)

FIG.%-Sketch map of Zones of Settlement on the northwestern slopes of the

Sierra Nevada de Santa Marta, Colombia. (GEOQ.Rezl., Oct., 1931; republished by
courtesy of the Am. Geog. Soc.)

protection from the northeast trade winds. Sixty feet of water is to be found
in this bay (Fig. S), and it is now one of the largest banana ports in the
world. The coast is rocky and rises fairly rapidly to 9000 feet in the San
Lorenzo Range. N7itliin 40 miles are the ice-clad peaks of the Sierra
Nevada, which rise to 18,000 feet or more. My chief aini in visiting Santa
LMartawas to find how the environment varied with elevation in a region so
near the Equator. The zones are suminarised in the tabular legend of
Figure 8. This table gives a good idea of the hinterland of the ton-n and of
its resources.
Tli~isthe settlement originated near the fine natural harbour as a
 32 TAYLOR-GENETIC U R B A N GEOGRAPHY [ Mar.

Spanish (nitpost to control the Spanish blain. According to E. R. Tirade i,

to~iartlathe end of the 16th centiiry the tun11 ahead? contained six of the
present rectangular streets and a churcli in what is now the southeast of the
towi. It \\-as however almost moribiuitl by 1882. when the railway to the
interior ivas begun. Ti itlay its prosperity depends almost \vIiolly on the
Innaiia trade, \vhich is most efficiently directed hy the American colony
ci miectetl with tlie United Fruit Coinpany of Boston.
- 1 s regards the to\vn itself it is built on the flat plain north CJf the little
hlanzaiiare5 River. The tcnvn. sheltered froiii the trade winds, is muggy
and hot. The natural vegetation on the coastal plain i s largely xerophytic
( P l i o t ~11 ) , I t is very dusty i n the frecluent dry spells. yet n u roads have
Downloaded by [New York University] at 06:16 15 February 2015

F' I I ( I T I I I I.-TIic tii\vii o f Santa Marta, perhaps the earliest settlement ciii the main-
land ~i .iriierica ( 1.W ) . Soticc. tlir arid vegrtaticiii (cactiis) iri the ioregrouiid. \Gel\-
loikiiis t o the south, with the Carihhean Sea t i ) tlie right.

lieen IiiiiIt intc.) the siirroiiiidiiig) IiiIIs, not even t o the coffee Iinriendas some
18 miles to the east. ivliicli are still reached only IIX mule tracks.
It is nc)t difficult tu distinguish that part of the tow1 \vhich is ~)urrl!.
Latin-.\mrrican frc ,111 that \vhicli has (1eveliq)ecl iiiainly as tlie result i li the
~~ ~

' FIG.O.-Llap of the city of Santa Marta iii Ciilomhia, based oil the city plat1 lO.?).
Reference : .\ Cathedral ; u Municipat Palace : c hIarket : n Church uf Sail Francisco ;
E (;overiiiIr'? Palace ; E' Cuartel : I I ' d i ~ l i i q ) 'Palace
~ : 1 Schools ; K I,aw Courts ; L H-IOS-
pita1 : 1 1 , s ,Q . Quintas. El Prado is the sulwrl) of the United Fruit Co. ( G r o ! ~Re?'., .
Oct.. IO3t : relluhIidietI hy&iurtehy of the .Am. Geog. SIX.)
Downloaded by [New York University] at 06:16 15 February 2015

19421
S A N T A MARTA, COLOMBIA
33
 34 TAYLOR--(;E N E T I C URBAN GEOGRAP €I Y [Mar.

recent cominercial growth of Santa Marta. There are four types of builtl-
ings, which together constitute practically tlie whole of the town. These are
arranged in definite zones, a i d are plotted in Figure 9. The foreign husi-
uess is chiefly in the west, and the local Iiusiness in the centre. The priorel-
people live chiefly in the east.
Oiie-storey buildings prevail save in the west of tlie town, a d only a €ew
of tlie larger puhlic oi- private buildings have three storeys. Santa Marta
is the capital of the large Delx~rtiiient of I2ilagdaleiia, and some of its
official buildings are quite pleasing in appearance. 111 the southwest of the
town are a number of d l a s i i i gardens, and they more closely reseiiible the
houses of northern peoples.
Korth o f tlie town has arisen a liari-ack-like sul)urli largely inhabited
Downloaded by [New York University] at 06:16 15 February 2015

by lalmrers on the \vharves. Very different is the suburb south of tlie

city where uiider beautiiul poinciana and acacia trees are the bungalows of
the Cnited Fruit Compaiiy.
IIow does Saiita Jlarta illustrate our discussiori on Possibilism ? The
fine harbour was 110 doubt a major factor determiiiiiig the original site,
\vhile the lack of any easily exploited resources led to the long stagnation OF
the town. I t will no doubt be stated that its present prosperity is largely
clue to the initiative of certain Bostonians. This is no doubt true ; but the
Bostonians were led to Santa Marta hq’ the unique environment ; in wliicli
n ~ i i i i e r ~ ice-fed
~is streams flow down to the Carilhean plains. Irrigation
is tlic key to tlie expaiision of the town in the last 40 years, and this is surely
a gift from llature ; developed of course by liumati initiative. The United
Fruit Company could build a railway anywhere along the Spanish Main.
They lvisely “chose” the one district where large-scale irrigation was
possible. The present writei- credits the expansion of a town based on 3
“HIihsoii’s choice” of this kind to the account of the determinists.

SECTION
B
CAKRERRA-THE CITY BUILT TO A PLAN
One of the most interesting human clusters is that of Canberra, the
capital of the Commonwealth of Australia. Most of the world’s chief
capitals have been fairly large towns before they hecame of great political
inqi3rtance. F o r instance, there \%-asa coiisiderahle population a t Ottawa
before it was chosen in 1858. 17‘ashington on the other hand seems to
have been of little or no importance before 1790, when Haniilton, it is said
in return for political favours by Jefferson, agreed to the latter’s desire to
place the capital on the Potoinac. However Canberra belongs to this ceii-
tury’s history, and every detail of its development is ascertainable. More-
over the writer had a fairly close personal association with the capital in its
earliest years, which may add to the interest of this brief record.
 19421 CANBERRA, AUSTRALIA 35

In 1900 it was decreed that the capital should be built in the state of
New South Wales. Rut pal-tly due to the jealousy of the other states it
was ruled that the city should not be less than 100 miles from Sydney (the
state capital). Since the western part of New South Wale!; is so hot and
dry that it is not suited for agriculture, it is clear that the new city would
develop in tlie rather elevated and wetter areas in the east of the state.
Various commissions visited suggested sites, and the first iiispection placed
Rombala ahead, with Yass and Orange second (Fig. 10). I n 1903 another
Downloaded by [New York University] at 06:16 15 February 2015

FIG.lO.-Map of eastern New South Wales to Frc.. ll.-Map of most of the

show the rival sites for the Capital. 1902-1905. The Federal Capital Territory in tlie
main isohyets are shown, and the regions Lvith over southeast of New South Wales.
40 inches are dotted. Note the dry area between The ruled areas are rather high
nalgety and Canberra. rI is Hillston. (Base-map and rugged. The small black
published by the author in Grog. Jill., April, 1914.) square near Acton shows the centre
of the new city. (First published
in the Grog. J d . , April, 1914.)

conimissioii favoured Albury, with Tuinut, Orange, and Armidale follow-

ing. Then Parliament voted, and chose Tuinut in 1903, and Dalgety in
1904. In 1908 at the filial discussion, Yass-Canberra was elected, through
the union of the parties supporting Yass, J>ake George, aiid Canberra, which
are all in the same region. The territory around Canlxm-a, coiiiprisiiig
900 square miles, was vested i n the Coiiiinon\vealtli on 1st January 1911.
 36 TAYLOR-GEh-ETIC b-RXAN G E O G R A P H Y [Mar.

Probably any of the ten places shown on tlie map would have been fairly
suitable. Canberra has however some advantages. I t lies close to the
route linking tlie two chief cities of Sydney and hIelbouriie. I t is 1900
feet above the sea, which gives a cooler climate thaii that of most of the
sites. I t is close to a long plateau (Biml)eri) which rises over 6000 feet,
and here the Cottei- river acts as a d u a l i l e collecting ground for the water
supply of the city.
Canberra is 011 the line joining the centre of population o f Australia (at
Hillston, H on Fig. 10) to the east coast. I t is not too far from the Federal
Downloaded by [New York University] at 06:16 15 February 2015

FIG.12.--i\Iap of the new city of Canberra, tlie capital of the Australian Common-
wealth. Note the Rfolonglo River running. west to the Murrumbidgee River. The
half dozen original houses, and the early roads (broken lines) are indicated. T h e
“cobwebs” in broken lines are part of the design, but have not yet been built. (From
the Official Map, 1933.)
 19421 CANBERRA-SITE A N D AXTECEDENTS 37

port at Jervis Bay, and yet it caiinot be shelled from the sea. O n the other
hand the rainfall is low (23 inches), since the site is at the north end of the
Cooma rain-shadow strip (Fig. 10). This however meant that the laid
was used primarily for grazing, aiid was not iii private hands. It is in tlie
faulted area of southeast Australia ; but earth-tremors, though not uncon-
mon in the district, have never been alarming.
The two first pastoralists settled in the district about 1823 ; and one of
these used the name Caiiberry for the district which surrounded his ranch,
which lie called Actoii. The other built the house named Duiitrooii on the
adjacent ranch to the east. The same squatter also erected the church at
Canberra in 1826. For a time Actoii was in the hands of the Rrasseys-a
Downloaded by [New York University] at 06:16 15 February 2015

well kiiowii naval family iii Eiiglaiid. Further west was Yarraluiiila, now
used by the Governor General. The whole Federal Territory was occupied
by a score of stations (ranches) mostly raising sheep, but iii the very
rugged southwest section heyoiid Tharwa cattle are grazed.
By 1909 there were a few tiny settlements serving these stations, such
as Canberra, Tliarwa, Hall, and Uriarra, some of which are sliowii on tlie
inap (Fig. 12). The total population was about 1500. The writer had
described the topography of the upper AIoloiiglo River as early as 1907 (in
his paper on Lake George), but did not visit the actual site until July 1910.
At this time tlie Federal Survey was housed in tents and in two fabric huts
just north of the Red Hill. Today the Prime Miiiister’s Lodge occupies
almost the same spot. (P.1cI.L. in Fig. 12.) A preliminary paper on tlie
topography of the Territory appeared in Decenilxr 1910 (illetcorol. Bzd-
letin 6) ; and I gave a full discussioii of the topography and history in the
Geogi-aphical Joitrnal (London) of 1914.
There were only two large houses on tlie city site in those early days.
One was Acton used by the Surveyor General, aiid tlie other was the
Rectory about a mile to the northeast. A little house near Ainslie was the
official post office (Photo 12). Duntrooii to the east sooii becaiiie the
nucleus of the Military College ; but it is still well outside the city, though
a flourishing suburb of the latter. The main road led from Queanbeyan to
Yass and ran just to the east of Ainslie. Here another road came up from
Tharwa in the south, while less important roads led to Actoii ant1 Yar-
raliiiula. All these are h w i i by brolceii lines 011 Figure 12.
After my return froiii Antarctica I was engaged 011 the geological survey
of the Territory. Our chief object was to find brick-clays aiid building stones,
which unfortunately are not altogether satisfactory i n the Territory itself.
By 1912 there were several wooden cottages near Acton for the surveyors
aiid clerks. A post office, drawing office, and bank had also been erected
iii the vicinity, and a large house (R. in Fig. 12) was built for the Adiiiitiis-
 3s TAYLOR-GENETIC LrRLIAN GEOGRAPHY [Mar.

i i
Downloaded by [New York University] at 06:16 15 February 2015

trator. Soon a SlllJI-t i-ail\va!. was rim i n f1-(~1:1(_)ueanl)eyaneight miles ti,

the east ; \vliile the Power Hi)use ( P.H. ) \vas 1)erliaps the most ainbitinus
Imilcliiig in those early (lays.
AIean\vliile plans ~ v c r einvitul h i - the futiirc cal)ital. a i i t l those i d ;L
Chicago architect (If’. u. Griffiri) \\.ere C h J S w . The actual street-idan
f i ) I l ~ i whis
~ design, \vliicli gave inucli 111-oiiiiurncetii the liical environmeiit.
The city has tlevelc ipetl in il 1)roatlflat plaiii I)et\veeii three cimical liills, \vliicli
risc respectively 400, 700. a n t l 800 feet al)ove the AIolo~iglo1ilain (at d m l t
1900 feet). The RIolonglo river w i n d 5 iii its flootl-silts acrciss this plain.
ruiiriing \vestn.ard tci joiii tlie IUur~-un:l)itlgreRiver (Fig. 12). The inaiii
yista is thc view nortlnvartls Iiet\veen F3lacli hlouiitain and Mci~iiitAilislie.
Here ~ ~ ~ l l l l l l ~ J.\venuel l \ ~ ~was ~ h thc axis ( i f the chief ccirrithr : and
~ ~ iiiatk
it leads nrrtliwartls ircim a Icnv hillucli, called Capital Hill, to the C i \ k
Centre :icross the AIolonglo River.
The colmeb pattern was adopted, wit11 two iliain cobwelx at Capital
Hill ant1 the Civic Centre. But other coInve1)s \Yere also laid out, most oi
wliich are still unly i i i tlie eiiil)ryo stage. Ho\vever in niy sketch map I incli-
cnte a tiumlier of these lesser c~iIi\velis: i.~. .. one niirtli\ve:;t of Ikl Hill aticl
thrce I)et\vet.ii the latter antl tlie railway station. Xiiotlier lies near Dun-
troon, while the Iarge5t cc~)l)\vehof all is quite outside the skrtch mal) to the
iiorth. This last is not yet 1)ortlered Iiy any houses. It will I)c ni)ticed
 19421 C A N BERRA-OCCUPI ED QUARTERS 39

that there is not much similarity I,etween the simple early road-pattern
(shown by heavy hroken lines) and the rather coinplicated cobwebs of today.
The actual arrangement of modern houses is indicated in Figure 12,
though every sinall street is not included. It will be seen that there are
three or four scattered quarters separated by M itle spaces where the streets
have not been developed yet. South of Capital Hill are the largest houses,
lvith the Prime 3Iinister's Lodge (P.M.L.) i n the north. Another quarter
served hy a cluster of shops and a theatre is to the east. Still further to the
east are the Railway Station, Power House and Printing Office with adjacent
houses. Parliment Hi m e and the chief Government buildings, are about a
mile north of these resitlential quarters (Photo 13).
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO 13.-Lookiiig southwest over Canberra. The view shows the Parliament
Buildings iii the right foreground. The long flat ridge of M m n t Rimheri (6264 feet)
appears in the background. (Courtesy, Governnient Printer, Sydney, N.S.W.)

The Civic Centre is three miles to the north across the river, and the
other maiii sections of the town lie between it and Moiint Ainslie. The
latter sections occupy the once eiiipty plain hetween the old Rectory and the
 40 TAYLOR-GENETIC URBAN GEOGRAPHY [Mar.

post office, where tlie writer iiivestigated the claims of certain water-
diviners way back in 1920.“
In the sketch map (Fig. 12) the streets in use are indicated, while m i l e
of them not yet laid out (near Red Hill a i d railway station) are suggested
hy broken lines. Presumably similar streets will in the future fill all the
empty portions of the iiiap, as well as the flat plain north of the Civic Centre.
A dozen small parks are scattered among tlie street blocks, but they are
omitted iii my sketch plan. The splendid plantations in the west of the city
are an attractive feature, atid a small hamlet has developed near the Forestry
Building to care for tlieiii.
Clearly we have here a departure from the plans of most cities of today.
The experiment of splitting up a siiiall town of soiiie 7000 people into
Downloaded by [New York University] at 06:16 15 February 2015

civic a i d political units is very interesting. The rather wide distance be-
tween tlie southern and northern clusters seems at present to waste a good
deal of citizens’ time. But in Canberra the designer is coininendably look-
ing forward to a city of three or forrr times the present size. It m i s t be
confessed that the choice of a cobweb plan supports the possi1)ilist’s thesis.

FIG.13.--Elock diagram of the region around Sydney, Australia, showing the

chief geological strata (1-4), and the three monoclinal warps. The sterile sandstone
is labelled 3. The area of shale soil ( 3 ) accounts for the closely-settled area in Figure
14. Note that practically all the road and rail traffic to the west passes through Lith-
gow. Inset is a sketch map to compare elevations between Sydney and Bathurst with
those between Newcastle and Cassilis.
“Water Divining” Griffith Taylor, Roq’rrl SOC.6’ictovia, (May, 1921). A map of
the empty site may be found of interest.
 19421 SYDNEY, NEW SOUTH WALES 41

The siting of the individual col)webs has however heen wisely influenced by
the relation of the river to the striking conical hills. Hence, as usual, in city
evolution, we must agree that nature and iiiaii both play important parts.

SECTION
C
SYDNEY-A METROPOLIS STRONGLY AFFECTED D T GEOLOGICAL CONTROLS

Having granted that the details of a city are man-made and lend small
support to the determinist, let us devote a little time to the broader aspects
of a city’s evolution. W e shall find that here the possibilist thesis meets
with much less encouragement. Sydiiey-the largest city in Australia, and
tlie second in tlie British Eiiipire-lias perhaps the most unusual uiiilaiid that
Downloaded by [New York University] at 06:16 15 February 2015

the writer has investigated. Its population is 1.3 millions, and in U.S.A. is
only exceeded by New York, Chicago, Philadelphia, and Detroit. I t offers
us a n interesting study, iii part because tlie first settlement was so recent as
1788. The writer lived in Sydney for over 20 years, during most of which
he was connected with the Uiiiversity there.
I n 1770 Captain Cook discovered the east coast of Australia, and he
spent soiiie days at Botany Bay (Fig. 13). Sir Joseph Banks was greatly
impressed by the new flora wliich he observed here; and later lie advised the
British authorities to make the first settlement a t Botany Bay. When tlie
First Fleet arrived in 1788, Captain Phillip was not much taken with the
wide sliallow bay or its sandy shores, and lie preferred Port Jackson about
12 iiiiles to tlie north (Photo 1 4 ) . This iiiagiiificent harbour had not been
visited hy Cook. W e may well consider for a few minutes the interaction
between iiiaii and environment in this locality.
The build of the Sydiiey district is relatively simple. A somewhat elon-
gated geological “hasin,” with its centre just west of Sydney, is built up
of three main formations. The foundation rocks are Silurian slates and
limestones (1 in Fig. 1 3 ) . Above these conies a thick series of Coal Mea-
sures of Pernio-Carboniferous age ( 2 ) . Two very valuable series of coal
seams are found iii these beds, of which tlie upper series lies about 3000 feet
below the city of Sydney. This coal outcrops on tlie coast at Newcastle to
tlie north, a t Coalcliff to the south, and also at Lithgow, soiiie 70 miles to the
west of Sydney. Above the Coal Measures is the deteriiiitiiiig factor in
human settlement arouiid Sydney. I t is a widespread layer of rather coayse
saizdstoize about 1000 feet thick, which covers an area of about 1200 square
miles. This layer of Trias sandstone (3 in Fig. 13) is usually bounded by
vertical cliffs, and these are suggested on its western edge in Fig. 13. Along
the coast both a t Sydney Harbour aud a t the mouth of the Hawkeshury
River, the coast consists of similar high sandstone cliffs. (Only the southern
third of this sandstone area is shown in Figure 13.) I t weathers to a sterile
 12 TAYLOR-GE K ETI C' V R H A N GEOGRAPHY I Mar.
Downloaded by [New York University] at 06:16 15 February 2015

PIIUTO I?.-Sydney Harhour. louliiiia t o the nurtheast. from the 1)ark around
Govertinient House. ~vliicliappears in tlie left foregrciund. The narrow outlet of tlie
Harliour to the Pacific is fvur miles away in the right I>ackground. ( Cuurtesy. Gov-
ernment Printer, Sycliiey, N.S.W.)

sandy s iil. \vliicli is covered \\.it11 heathy plants (ir scrulhy trees, and sup-
ports i i ( i crops o f iiiil)ort:uice. ;il,o\-e this Hawkesbury Sandstone is a layer
of Trias sliale (-I), \vliicli gives rise to clay SCJilS of much greater agricultural
value.
I'rolJaIdy ii these siiriace strata \\-ere relatively level and uiidisturbed,
far better conditions for cI(ise .settIenieiit \ \ ( ~ ~ i l thl v e ari,seii, tlirough the
deposition of silts, Iiiiiiiiis, etc. Hiit quite lately in geological time the ~ i i a r -
gins of this irregular basin \\-ere warped ulnvard. To the riorth of Sytlney
is a fairly steep \\-arp aiiiciuiitiiig to SO0 Fret, t o tlie sozrtlz the land was raised
hy a gentle incline tlJ a Iiciglit (of 1000 feet. \vIiile t(J the wcsf-at a distance
of 35 miles--a very marked hinge-line tleveloped. T o the west of this line
the surface was elevated to a height o f ,3000 (ur 1000 feet iii a remarkal~le
"iiiclnc~icliiie." As the land rose. the Nepean Kivcr was forced to fhiw nlotig
the remarkal,ltr series ( i f short gorgcs and alternating flats which I have
tlescril)ed elsewhere. (These latter arc omitted in the sketch Fig. 13.)
A s a result of these uplifts, especially i i i the west, the top(-igraphyaround
Sydney is much tlissectrtl. Ititleetl near Katoomba are canyons 2000 feet
 19421 SYDNEY-DISTRIBUTION OF SETTLEMENT 43

deep, of which soiiie (like the Grose Valley, Fig. 13) elid to the west in giant
cliffs (Photo 15). This dissected topc)grapliy lias always Iilocked con~iiiuni-
Downloaded by [New York University] at 06:16 15 February 2015

PtroTo 15.-This wide canyon near Eatoomba is about 8 miles across and 2000 feet
deep. It is eroded in thc Triassic sandstones of the Blue Plateau. This canyon has a
b i attle-neck wtlet-gorge only a few hundred feet wide, where it enters the Nepeaii
Rimr in the plains. (Courtesy, Governmtwt Printer. Sytlney. N .S.W.)

cations, :itid indeed nut for a quarter of a century cciulcl the settlers fitid a
u ay across the santlstone plateau. Hou ever the centre of the original Insin
\vab nt)t elevated ; 50 that lirre the upper layer o f h a l e s was preserved, while
the soft shales were w" remc rvecl froin the up-warped inarginal surfaces.
\Z'e iiow have the liey to the remarkahle unilantl o f the city of Sydney.
The distrilxitioii of popiilatioii is given in Fig. 14. If we draw a circle of
50-mile radius almit Sydney (with its 1.3 million inhabitants) we find one
of the most singular dispositions of circuiii-metrc)politan population in the
worhl. Let us suppose that b e survey this Iielt from an aeroplane. On the
iiortli oiir flight leaves the narrow coastal plain at Wyee. Thereafter for
80 miles \ve crabs only one good road. The next 20 miles crosses Bell's
Road and the tourist ceiitre of Katooml)n Proceeding south for 40 miles
we notice the single settleiiie~itcrf the Yerantlerie miiies, before we reach the
 44 TAYLOR-GENETIC URBAN GEOGRAPHY [Mar.
Downloaded by [New York University] at 06:16 15 February 2015

FIG.14.-The distribution of settlement around Sydney. The region of city streets

is shown black. The region where centres are less than about four miles apart hy close
ivfliny. (S.B. Closely settled areas begin again somewhat to the west and north of
this map.) (From R o y a l Soc., Nrzo Soztfh LVtrles, 1923.)

railway line at Balmoral. Another 20 miles of uninhabited water-reserves

brings us to tlie narrow coastal plain near Kiama, and here again a closely
settled region is observed. Tli~is,excluding tlie shore, during a flight of over
150 miles we have only crossed two narrow belts of settlement (at Katooinba
and Balmoral) while circumiiavigatitig the largest city in Australia.
It must of course lie realised that, after we have passed to the west of the
sterile Hawkesbury Sandstone, tlie normal density of a farming population
is soon reached again. For the most part this hetter country lies outside the
area of Figure 13, hut a little of it shows near Litligow.
It may be noted that though Sydney is by far tlie largest ton-n Tvvitli
probably the best harbour in Australia, yet it has these grave disadrantages
of restricted farmlands and difficult approaches to the hinterland. At New-
castle 70 miles to tlie north, a large river, the Hunter, enters the sea after
flowing through much better agricultural land. A quite adequate harbour,
in part artificial, is to be found at Newcastle. Here also outcrop tlie liest
coal seams in the southern lieinisphere. Beliind Newcastle is the Cassilis
Gate ; a broad low corridor of access to the interior, and far superior to the
Litligoi\;-Batliurst route, as is obvious from tlie inset map in Figure 13. In
the latter case the main Western Railway has to climb up to 3300 feet or
more on its way to the west, while the Cassilis Gate is little more than 1000
feet above sea level. Yet there is still no railway through Cassilis; and
 19421 TORONTO, ONTARIO 45

Newcastle with all its advaiitages has only a population of 105,000. If

Phillip liad decided to fouiid his first settleiiient at Newcastle, one wonders
if Sydney might not have been nearly as empty today a s the siinilar harbour
of Broken Rap (Fig. 13). Here then is an example of man’s choice of a city
site with s w i e iiotahle disadvantages, which 111 part offset the spleiidid
harbour. The writer will not atteiiipt to coinpare the respective total attrac-
tions of the two sites of Sydney aiicl Newcastle, but the topic is of more than
local interest. The aiiswer may he clearer iii a huiidred years’ time.

SECTION
D
TORONTO-SrAGES I N T H E DEVELOPMENT O F A LARGE CITY
Downloaded by [New York University] at 06:16 15 February 2015

Let us study the way in which a large city, such a s Toronto, has evolved
in about 150 years. Its topography is sliowii in Figure 15 ; while the main

FIG.15.-Block diagram of the City of Toronto, showing the topographic features which
have affected the evolution of the city. Inset is a diagram showing the locality at the Lake
Iroquois period. The fine lines labelled 1889 and 1932 show the edges of the streets with close-
set houses at those dates. \\ 1-\r3 is proposed Woodbine Avenue ; “Wal.” is Walmer Road.
(From Caw. JuZ. Ecoiz., 1936; with additions.)
 36 TAYLOR-GENETIC U R B A N GEOGRAPHY [ IZI;n-.
Downloaded by [New York University] at 06:16 15 February 2015

~~~~ ~ ~ ~
~~~ ~~~ ~ ~~

FIG.16.-Stage-diagram to show the evolution of the city of Toronto. The four maps are
to the same scale, the earliest at the bottom. The various rulings indicate the various elements
of the city Dattern. 1. 2. 3 . 4. are the four types of city residences, varying from mansions (1)
to the poorest types ( 4 ) . The zones are. of course, somewhat generalised. The “coinmercial
core” includes the large Department Stores.
 19421 TORONTO-STAGES I N SETTLEMENT 47

stages of development are shown in Figure 16, where only the salient cliarac-
teristics are charted. I n a large city we need to know the position of the
coniniercial core, the industrial areas, and the general characters of the resi-
dential quarters. The latter may be fourflz class ( H d ) , siiiall inconvenient
(or decayed) houses without gardens ; third class ( I k ) , better houses but
still without niucli space or convenience ; second class ( H b ) , iiiodern two-
or three-storey brick houses with fair gardens; or lastly first class (Ha)
with rather large gardens for a town, and of the type that at times might be
called mansions. These types of residences are numbered 1, 2, 3 , and 4 in
the maps for 1885 and 1910 in Figure 16.
If we ignore the other details we see that there is a constant change in
Downloaded by [New York University] at 06:16 15 February 2015

the position of these zoiies. This change takes place in two ways. The best
class houses of one decade become the second class houses of a later decade.
But this change in zonal distribution is complicated by the building of new
suburbs of various classes in the iiiargiiial hitherto empty regions. There is
therefore no very close resemblance between the developiiieiit of a city and
that of an animal. But it is something like the way a young tree grows. Its
trunk and branches increase by a sort of expansion, while quite new cliarac-
ters such a s flowers aiid fruit appear a s it reaches maturity.
Tlie lower diagram in Figure 16 sliows Toronto in 1818. The original
grid of nine streets was laid out about 1790, on well-drained grouiid a t the
head of the enclosed bay, near tlie Don outlet ; where also the best protection
from naval attack was presuinably to be gained. The junction of King and
George streets was tlie early centre o€ the city. By 1818 siiiall industries
such as tanning a i d brewing were scattered near the original settlement.
A few larger houses were lmilt on the margins of the town, hut all but a few
of the dwellings were still built of wood. The population was about 1200
a t this time, and tlie built-in area covered about half a square mile. As usual
in such a young organism it is not possible to distinguish the zones, which
will only differentiate at a later date.
By 1842 tlie town liad about 20,000 inhabitants. The centre liad shifted
to tlie lower end of Yotige Street, the main corridor to the interior. Swamps
near tlie foot of Yonge Street teiided a t first to isolate the old from tlie new
portion of the town, but these were filled in by degrees ; so that hy 1842 this
swampland was becoming the commercial core of tlie city. The houses were
still small for the most part, but finer residences were being built on tlie
northern margins. One of these was tlie Grange, which is now part of the
City A r t Gallery. Tlie built-in area (diagonal ruling) now covered about
two square miles.
I n 1885 the town contained about 120,000 citizens, and extended over
most of the City Plain, between the lake and the 40-foot cliff of the old Lake
Iroquois Shore-line. Tlie coiiiinercial core has moved a trifle north, but
 48 TAYLOR-GEKETIC URBAN GEOGRAPHY [Mar.

still keeps to the Yonge Street axis. Tlie houses are now numerous enough
to sliow a fairly clear soiiirig, ranging from tlie poorest houses ( 4 or H d ) up
to near-mansions ( 1 or H a ) . Tlie underprivileged live near the lieart of
the city, in old houses for the most part. The wealthy are building large
residences on the margiiis. where parks (as in tlie west) or deep ravines (as
in the east) promise security from further encroachment. The built-in area
covers approximately eight square miles. The city is definitely growing to
tlie northwest, mainly because tlie deep wide valley of the Don bars ready
access to the east. Tlie little streani-valleys also guide the expansion i n the
same direction. I t is still not easy to differentiate between the commercial
aiid iiidustrial areas of the city. In a later section (11. 59) we shall label
Downloaded by [New York University] at 06:16 15 February 2015

such a stage of developinent as Early !Mature.

In tlie modern city of 1940-which contains nearly 900,000 inhabitants,
if adjacent satellite.; are inclutled-diere has I ~ e nan enoriiious expansion
of houses i n every direction, h i t especially to the north and northwest. The
old shore-line ceased to be the northern boundary almut 1890. A great de-
velopment of industrial plants has taken place, especially along the main
railways west of the city, aiid on the filled-in lake areas to the southeast of
the city. Tlie coininercial core has broadened somewhat, Ijut is still clus-
tered about Yoiigc, Queen, King and Bay streets. Here also are the large
department stores, which have not been differentiated in my discussion.
In the last half century the tliird-class houses liave hecome the poor-class
houses of 1940. The earlier first and second-class houses have been heniiiied
in by the growing city, riddled with sniall shops ant1 with small industrial
establishments. They have become the third-class houses ( 3 ) of today. To
keep pace with the immense groudi of the city, now seven times as ~ O ~ L I ~ O L I S
as in 188.5, vast areas of thirtl-class houses cover most of the City Plain : hut
naturally many of the better houses (2) hold their own in the marginal por-
tions of zone 3. The iiiotlerii zone 2 has developed mainly along Yonge
Street, the main corridor to the north, especially near the h i e west-east
avenue called St. Clair. 111 the extreme west near the Huniber, and tlie
extreme east towards Scarborough Cliffs, there are other clusters of second-
class houses.
Tlie positions of the four first-class suburbs (1, or H a ) are interesting.
Partly they occupy the margins of ravines, as near La~vrencePark (L.P.) in
the north, Rosedale (K.D.) i n the east, aiid Humberside in the west. But
the development of Forest Hill is perhaps due to the times of prosperity
about twenty years ago. Tlie high rolling hills of till, herc about 300 feet
above tlie lake, were beyond tlie city a t that date, and were rapidly covered
by large and expensive mansions. Possibly no similar developiiient will
 19421 STAGES I N URBAN DEVELOPMENT 49

occur again, since both tlie children in families and domestic servants are
now rather rapidly declining in numbers.
Teizfative “ages” of a city
1 have labelled tlie four stages shown in Figure 16 with the names so
familiar to us in regard to an evolving landscape. Most progressive cities
of the occidental type go through similar stages. I n tlie earliest stage there
is no clear differentiation between industrial, commercial, or residential
areas. T h e first tendency is for the bigger houses to develop near tlie
margins, where gardens and privacy are available. We see Toronto in
this infantile stage up to 1818.
I n the next stage there is a fairly clear segregatiou of aii extensive coiii-
Downloaded by [New York University] at 06:16 15 February 2015

iiiercial quarter towards the centre of the town, including a number of streets
without residences. But shops, offices and small industries are still rather
mixed. A definite zone of better-class houses fringes the town. Tlie poorer
folk live near the centre ; in many cases taking over tlie better houses of tlie
former stage, but also building new small houses in any empty spaces. This
may be called tlie juvenile stage, aiid is illustrated around 1842 in Toronto.
Tlie beginning of maturity is sliown by a definite differentiation of tlie
residences. The various types are displaced outward as the years move on,
though naturally examples of tlie early houses survive in tlie expaiitling
zones. I n 1885 tlie four classes of residences mentioned earlier were begin-
ning to become evident. Poor quarters (4) clustered around the centre of
tlie town. Tlie best residences were on the iiiargiiis, but were not iiiiinerous
enough to form a connected zone. High Park in the west helped to deter-
mine the zone. The ravines a t Kosedale attracted another area of good
houses. This stage is adolescejit or early mature.
I n tlie mature stage of Toronto we see that the concentration of industrial
areas along the railways and on the made ground is tlie chief change in tlie
conditions. Toronto is still growing, but tlie main axis of growth (i.e., tlie
central line of the new industrial zone) is iiow along the railway to the
northwest. There is a separation of tlie industrial city from the residential
city, aiid the city is iiiatwe.
Perhaps when a definite zoning is carried out according to a laid-down
plan (such as is indicated in Canberra) a city may be said to reach its zenith
or climax ; in geographical language it is late 77zatmre. The senile condition
of a city is not indicated in Toronto. Such towns as Pekin (Fig. 5) or Nan-
king, with vast areas originally covered with houses but now heaps of ruins,
may perhaps be described as partly senile. But tlie complexity of a large
and ancient city is like the coniplexity of a long river like the Nile. Parts
of tlie Nile are juvenile, others mature, others senile, depending on external
factors developing during tlie life of the river.
 50 TAYLOR-GENETIC URBAN GEOGRAPIIY [Mar.

aiid the city’s growth

Eizvi~o~ineiit
A city is obviously an artificial growth. Nature produces nothing resem-
bling a city, and so at first glance would seem to have no say in the develop-
ment of a city. I have discussed this aspect of the pro1)leni as regards
Toronto in several publications.”’ Here therefore I will only introduce one
cfiagram (Fig. 15), and point out a few examples where nature’s lay-out
might \veil have been considered in deciding man’s design.
The Toronto environment consists of a topography built up wholly of
glacial debris, either till or sands. These are 300 feet thick in the upper
part of the town, and about 60 feet thick in the lower part. Rock outcrops
are absent everyahere. except in tlie lied of the Hmnber. The City Plain ex-
Downloaded by [New York University] at 06:16 15 February 2015

tends from Lake Ontario (246 feet) up to the Iroquois shoreline (425 feet).
Here a sharp slope, alniost a cliff 40 feet or more in height, (which was cut
out by the waves of former Lake Iroquois), separates the lower from the
upper town. The Canadian Pacific Iiail\vay runs along the foot of the
Ircquois shoreline and definitely blocks road conimuiiicatioii with the upper
suburbs.
Tlie chief natural features of the toLvn, however, are the ravines, which
have been eroded in the till by the drainage from the foi-iner ice-sheet. These
are 50 to 100 feet deep, and in general run from northwest to southeast (Fig.
1.5). Tlie civic fathers followed the usual chessboard lay-out of the streets,
with their axes near the cardinal directions. Hence the roads have been
made without the slightest reference to tlie ravines, and without much regard
for the main objective of the citizens’ journeys (which was to reach the
centre of the city). Some of the most expensive city-viaducts in the world
have been built to cross the Don and Huniber valleys. These were inevitable,
but others over the smaller ravines need not have been bui!t if the roads hat1
been built aloiig the divides between the ravines. In the west part of tlie
town such roads n-ould have tlie added advantage that they led directly to
the centre of the city. F o r instance, Vauglian Road running diagonally
betwcciz ravines is a inuch more sensibly designed road than Spaclina Road
with its huge viaduct.
In Toronto, as in many other fairly modern cities, the chessboard was
laid out on the surveyor’s chart. and often the roads mere nanietl before
construction. Thus we find a continuous Walnier Road in the middle of
the city map (see Wal. on Fig. 1.5), but actually it is still in three isolated
pieces. These sections are separated from each other hy the Iroquois cliff,
or by deep ravines, or by the railway.
The most ainusing example is Jfroodbine Avenue (Wl, 1172, JV3 on
~

l o“Topographic Control in tlle Toronto Regon,” Griffith Taylor. Cnrt. J d . Erori.,

(Nov., 1936).
 19421 T H E “ZONES A N D STRATA” CONCEPT 51

Fig. 15) which lias a gap three miles long in its continuity from W2 to W3.
I t seems hardly likely that tlie whole of the Little Don valley will be filled
up to enable tlie surveyor’s design to be maintained ! Yet this is merely a
large-scale example (of which we see many smaller illustrations) of the
absurdity of ignoring tlie environment in planning the development of a
large city. Lack of space forbids iiiy enumerating many other examples of
man’s unsatisfactory choice of plan in this connection.
The writer would like to enter a plea for tlie use of the block diagram
(as in Fig. 15) to illustrate the setting of a city. Its history may be so easily
suiiiinarized therefrom, as follows :
T h e first houses in 1793 were built near tlie black square in the south-
Downloaded by [New York University] at 06:16 15 February 2015

east. This is the “nucleus” of tlie city. By 1818 the little town had spread
along tlie shoreline. T w o isopleths, indicatiiig the outer limits of built-in
streets in 1889 a i d 1932, show the expansion of the city a t two later periods.
The large valleys of tlie Huiiilier and tlie Don, tlie small glacial-water ravines
and the Iroquois shoreline are clearly represented. This illustration should
be used in conjunction with the stage-diagraiii (Fig. 16).

SECTION
E
T H E “ZONES A N D STRATA’’ CONCEPT APPLIED TO CITIES ; TORONTO,
PORT CREDIT A N D W H I T B Y

I n the case of large human nioveiiients-such as racial migratioiis-the

use of the zones and strata technique enables us to deduce tlie cradle of human
evolution a i d the order of development of tlie races.ll But it is also of coil-
siderable value in connection with tlie much smaller agglomerations of man
which we call villages ant1 cities. Here also we are concerned with clusters
of folk about the original nucleus, about which the town gradually expands.
What value has this technique in city evolution? Perhaps in connection
with house-types we see the process most clearly. The first houses at
Toronto were log Iziifs, of which one is actually still preserved in tlie Exhibi-
tion Grounds. This was origi nally erected where Queen Street crosses tlie
Don River (Photo 16). The nearest log hut, in m e today iiiucli a s it was
originally, is (as far a s I know) in Queensville, 35 iiiiles north of Toronto.
Thus the first stage of house lias now been pushed out for 30 niiles or so
from its original position. Tlie second type was the small fmnze house built
of horizontal planks. I recently photographed one of these houses in the
heart of tlie city in Queen Street opposite Osgoode Hall (Photo 1 7 ) . I t has
now been pulled down, hut plenty of specimens still remain in corners of the
city. They are now found fairly numerously in the little towns near Toronto ;
See the writer’s Eiwirorztizeaf, Race aiid Miqratioiz (Chap. 21).
 52 TAYLOR-GEKETIC U R B A N GEOGRAPHY 1 Mar.

P H o m 16.--T\vo log-cabins which illustrate tlie application of the “Zones and Strata”
Downloaded by [New York University] at 06:16 15 February 2015

tc.chiiique to the evulution o f T#mmto. The cahin I111 the kit was built about 1x00 right in the
early village uf Toronto. Today. the occupied cahin OII the right ( a t Queeiisville. 35 miles from
Toronto) s l i ~ t v s111mvfar this tylw o f I)uildit~phas hem displaced frcim the ceiitrc of the city.

Pkioro 17.-:\ frame hrriltliilg. over 100 years old. long siirviving in the main
street of Toronto oppusite the La\\. Courts. Today such houses are displaced more than
7 miles atvay frcitii the city.

aiitl \ve may say that the!. 1iai.e been tlisl)ersetl to ;I (listatice of 7 o r 8 niiles
from the nucleus (Fig. 17).
The fnr~rrilroirsc-with a 1m-n perlial)s coiiivs nest. n‘one is left iii the city
as far as I know : 1mt they still occur on the margin, at R distance of almut
. imiles, as for instance ileal- \2‘ilson .\venue. T h e f(:iurtli type is the two-
storey brick Iiiiuse often wit11 n snc~\\--sIietltling galde in fri)iit. Tliese a r e
still coiiiiiiu~iin the third- aiid f(ourth-class house^ in tlie city. They are of
 19421 TYPES O F STRUCTURES 53
Downloaded by [New York University] at 06:16 15 February 2015

FIG.17.-A block diagram illustrating the “Zones and Strata” technique applied to
the evolution of Toronto. The first houses were log-cabins now displaced 35 miles
away from the city; but an early specimen still remains in the city. So also the later
types of frame-houses, small brick houses, mansions, etc. have been displaced succes-
sively from the nucleus of the city.

course almost universal throughout Lower Ontario, being the usual type of
farmhouse built during the last half century.
T h e remaining types are the mansions and sky-scrapers. The Grange
was built in 1817 only a mile from the Lake. I t is a large ? J X U Z S ~ Ostill
~ ~ in
use as part of the Art Gallery. Today houses of this type and culture are
not fouiid nearer than Kosedale or Forest Hill, i.c., about 4 miles north.
Finally in the busiest part of the city we find the latest type of building, the
sky-scrapfr, which is almost wholly found in the vicinity of Queen, Bay, and
Yonge Streets. They of course coiistitute the most striking buildings in the
commercial core.
The zones are therefore fairly definite today. If we assume that the few
relics of older types still remaining are fragments of ancient “strata,” most
of which have been “buried” (it., removed), then the evidence given above
shows how the technique can be applied to city evolution. In other words
all the types originated near the heart of the niotlern city (near the “nu-
cleus”) ; hut tlie older tlie type the further it has been displaced to the mar-
gins as tlie city evolves. Conversely from the distrihution of these zones
we can deduce the way in which the houses have evo1ved.12
The data can be suinmarized in a table as follows, with the oldest types
of building at tlie foot of the table.
12 T h e writer first used this technique for Chicago ; see his paper “Geograpliy. the

Correlative Science.” Can. JuZ. Ecoiz. (1935), p. 542.

 54 TAYLOR-GENETIC U R B A N GEOGRAPHY 1Mar.

Order 1 Type I
~~
First site Present position
____~ -
Displacement
from centre
6. Sky-scrapers Centre of city Centre of city ’ ( h r e a 1 mile wide)
5. Mansions do. Margin of city 3 or 4 niiles
4. Early brick houses do. Throughout province ?
3. Farmhouses do. ? do. 5 miles
2. Frame houses do. Especially in pioneer
areas 7 miles
1. ’ Log cabins do. Obsolete 35 miles

The circuiiispicc’3 coiicejt illizstratctf by Port Civdit and Whitby

Downloaded by [New York University] at 06:16 15 February 2015

Darwin, while studying coral reefs, reali sed that amid the many examples
in the Pacific there were likely to be reefs in all stages of evolution. 111
effect he said “ L O Caround
J~ you, and yo11 will lie able to reconstruct the
evolution of the coral atoll.” The epitaph of IT-reii tlie architect in Saint
Paul’s in a sense contains the same idea, “Si reqziiris moii uiiieiitairi, circum-
spice.” IVe can usefully apply the same concept of “looking around you” to
city evolution. The village is father to the city; hence if n e study a village
in the same enviro~iiiienta s the city, we shall learn soiiiethiiig as to an early
stage in the evolutioii of the city.
I have made a number of surveys of towns and villages in southern
Ontario, and I think it will be of interest to discuss two of these. Port Credit
is a little town on the lake 13 miles west of Toronto. I t has about 2000
inhabitants, and in some ways resembles an early stage of Toronto, when
the latter contained about the same number of folk. (It is not of much sig-
nificance that modern features such as electricity in Port Credit had no
counterpart in the early stages of Toronto.) My second illustration is the
town of \;2’liithy (SO00 inhabitants), which has grown up about 27 miles
east of Toronto with iiiuch the same topography and hinterland as the other
two lake settlements.
Port Credit was first settled in 1804 a t tlie mouth of the Credit Kiver
(Fig. 18), and until 184.5 the village was confined to the western bank.
-4hiost all the houses were wooden, and were iiiostly two-storey gabled
residences, of which several still survive. The next few decades were quite
prosperous, and the little town had a considerable trade with U.S.A. in grain
and timber (Photo IS). There ivas a busy port with quays, harbour works
and a lighthouse. But tariff laws ruined the American grain trade, while
13 If you wish to pronounce “Circuinspice” ;
Don’t rhyme it with “Mice” . . . hut with “Mickey”!
 19421 POST C-REDIT, ONTARIO 55

all the titnber was exhausted in the hinterland during the seventies. The
storms around 1878 demolislietl the I~arbourworks, and they were never
rehuilt.
T h e railway station was placed on tlie east side of the estuary in 1855,
and this led to the development of a new town nearer to the station, in which
Downloaded by [New York University] at 06:16 15 February 2015

FIG.18 ( k f t ) . - A functional plan of the little town of Port Credit west of Toronto. T w o
contours are charted. The houses are in four classes ; Ha. Hb, Hc, and H d ; tlie latter are mostly
shacks.
FIG.18 (ri,qht).-A plan of Port Credit reconstructed from the f o r n d a shown at the top
of the figure. I t is to he compared with the actual plan on the left.

PHOTO 18 (Icft).-The site of the former flourishing timber and grain port at the mouth of
the Credit River. Today, though Port Credit is a good deal bigger, there is hardly a vestige left
of the wharves and breakwater of 1875. The view is to the west.
PHOTO 18 (right).-Oiie of the large frame houses which were typical of Port Credit about
a century ago. Very few are now left.
 FIG. 19.-A functional
plan of the little town of
Whithy, 23 milcs to the
east of Toronto, and in a
similar environment. The
house-zones 1, 2, 3, 4,
Downloaded by [New York University] at 06:16 15 February 2015

correspond to the zones

Ha, Hh, Hc, Hd, in Fig-
ure 18. Inset is a recon-
1 struction of the plan. de-
rived from the formula
I , .
given i n the text. S is
1 school. Ch is Church, Sn
is station.
 19421 I V H I T B Y , ONTARIO 57
houses were built much inore generally of brick. The population was about
450 in 1877, when tlie town contained three churches aud three hotels;
which is the same as today with a population almost 2000.
I n 1900 the eaytern portion equalled the western in area, but since that
date there has been a great expansion to the east, as Figure 18 shows. Thus
Port Credit developed in three stages ; the old wooden town in tlie west till
1855 or so ; the modern brick aiid wood town between the station arid the
little harbour till 1930 ; and after that date tlie newer portion right in tlie east.
There are two large industrial plants, each giving work to about 100
employees. The Starch Coinpaiiy cotnineiiced about 1889, and the large
Oil Refinery in 1933. Today the town contains two schools and a High
Downloaded by [New York University] at 06:16 15 February 2015

School, two banks aid a large brick post office. The two rows of shops
shown 011 the iiiain street comprise 3 stores, 4 cafes, 3 realtors, 3 lnrbers,
2 butchers, a drug store, and a picture house.
The town pattern is endy mature, since the older wooden houses were
long ago replaced by brick near the shopping centre. The cheap wooden
houses ( H d ) have grown tip on the margins, partly in response to the two
large industrial plants. The very slow growth of the town has prevented the
change of early homes ( H c ) into the near-slums ( H d in part) as in Toronto.
The group of large brick houses (Ha) in the southeast is not a natural out-
growth of Port Credit, but a sort of forest-suburb of Toronto.
I t seems likely that the oldest western part of the town, consisting largely
of wooden houses, gives us some idea of Toronto about 1840, though the
latter was of course far larger (Photo 18). T h e addition of the newer east-
ern portion of the town with its nunierous two-storey brick houses affords
a picture which to some extent resembles Toronto in the seventies, before
the development of special industrial sections in the city. The large houses
(Ha) in the southeast, each surrounded with a large garden and often by
small forest trees, are really outlying parts of Toronto ; for their owners have
in general not much cultural connection with Port Credit. However there
are still inany acres of farnilatid separating the little town from the metropo-
lis. Not yet has it joined Long Branch and ILliiiiico a i d become a mere
satellite of Toronto.
A second example of a small town on the lake-shore near Toronto will
show us the normal development of a town coiisiderably larger than Port
Credit. I have chosen Whitby, which is 22 miles east of tlie metropolis.
Its site is much the same as at Port Credit, save that there is no river there,
but only a sinall shallow bay which has been converted into a harbour by a
breakwater across tlie mouth. The main cross-road follows along one of the
usual broad north-south ridges of this coast. The 300-foot contour is shown
011 Figure 19. The swamps near the harbour have prevented house develop-
ment near the lake. As a t Port Credit, the days of shipping grain and timber
have passed. Now oil is the chief import, and is stored in €our tanks which
 58 TAYLOR-GENETIC URBAN G E O G R A P H Y 1Mar.
take the place of bygone elevators. A suburb south of the Canadian National
railway survives from these earlier days. The total poptilation is about 4500.
\$‘hithy today has developed at the cross-roads, where tlie Toronto-
Kingston corridor crosses the north-south road to the little port. The inipor-
tance of Whitb~7as a port has almost vanished, hut a number of sinall fac-
tories, dealing with brass-work, tanning, Ihiikets, and canning, are adding
to the vitality of the little agricultural toaii. The factories are not yet segre-
gated, indeed several of them are not even on the various railways which
meet at Whitby.
The pattern of the town shows the usual inner mile of shops at the cross-
roads together with the post office and T o ~ v nHall. Outside this is the zone
of larger houses, nearly all of brick (2 in Fig. 19), and here are found the
Downloaded by [New York University] at 06:16 15 February 2015

chief churches. Beyond these zones tlie houses are on the whole built of
wood, the nearer being usually two-storey and tlie distant houses one-storey.
However a few larger houses are of course to be foimd in these zones also.
A large Ladies’ College is isolated on a ridge to the east of tlie town.
Although the town is twice as large as Port Credit it is not in quite such
an advanced stage of urhan development, mainly because the factories are
smaller. The town may be described as in the atlolcsceizt stage, for there is
no itidustrial zone and iio zone of larger residences. Owing to the vicinity of
Toronto, which tends to attract so much of the local trade, it does not seein
likely that \Vhitby will soon advance much beyond this stage of development.

F
SECTION
THE CLASSIFlCATION O F TOWNS

There are innuiiierable types of towns scattered over the face of the
earth, but it seeins possible to clarify our ideas a little by “looking around”
for stages in evolution as suggested earlier. The earliest settlements and
towns seem to have developed in Mesopotamia or thereabouts. W e can
undoubtedly still find examples which differ little from those evolved maiiy
thousand years ago. They de\-eloped in turn into the crowded oriental towns
of today. Perhaps one useful distinction is that between the oriental (or at
any rate non-European) town and the occidental town of Europe or North
America. This latter had its fairly R-ell-defined evolutioii, which Geddes and
Mumford have e ~ a m i n e d . ’ ~\Ye may use their terms, which I have briefly
defined in parentheses. Eotechnic (walled) towns, Baroque (palace) towns,
Paleotechnic (industrial) towns, Xeotechnic (planned) towns, Biotechnic
(regional) towns, can all still be studied if we look far enough afield.
In accord with the Zones and Strata Concept the more primitive towns
14 Patrick Geddes, Cities &IEvolutiorz, London, 1915.
Lewis Alumford, Cailtzi~eof Citirs, Kew York, 1938.
 19421 CLASSIFICATION O F TOWiYS 59

will be fouiid near the margins, the iiiore advanced near tlie centre of stiiiiu-
lus. Accordiiigly near New York aiid Londoii we see the latest types,
garden cities such as Radburn ( N . J.) aiid Letcliwortli near Loiitlon. These
are perhaps Neotrclziizc iii character. But the great majority of our occi-
dental towns are Palcotcclziiic. Still, in quiet districts such as Carlsruhe or
Versailles, we may gain some idea of the Baroqzie town of 200 years ago.
Still further afield Old Carcassonne and Aigues lLlortes in the south of
France have changed little from the Eotechiiic style. In Africa the oriental
town still remains very little altered by a thousand years of progress else-
where.
In the present address I aiii almost entirely concerned with the indus-
Downloaded by [New York University] at 06:16 15 February 2015

trial (Paleotechnic) type of town. It is tlie subdivisions of this type with

which we are immediately concerned. I have already suggested that most
cities of our age and culture pass through a iiumber of similar stages. It
should be the duty of the urban geographer to integrate his data in this
research, much as W. M. Davis did in his classic studies of the evolution of
landscapes. But our probleiii will be inore difficult, since our agent-deter-
mining the evolution of cities-is iizan the irrational, while the agent in his
problem was ever-logical Arature.
The matter may well be set out in a tahle a5 follows :
; SUBDIVISIONS
CLASSESOF TOWNS OF PALEOTECHNIC
Town-s

Evolution of towns Occidental or Oriental.

Occidental towns Eotechiiic, Baroque, Paleotechnic, Neotechnic, Biotechnic.
Paleotechnic stages Infantile, Juvenile, Adolescent, Mature, Senile.

Infantile towns ( I n ) Haphazard distribution of houses and shops, no factories.

Juvenile towns (Ju) Differentiation into zones (of houses and shops) starts.
Adolescent (Ad) Scattered factories, 110 definite zone of ‘Ha’ houses.
Early Mature ( E r ) Residence zones fairly defined, no segregation of factories.
Mature (Ma) Four zones of houses, separate commercial and industrial areas.
Late Mature (La) Indications of advance to biotechnic development.
Senile (Se) Large areas of town abandoned, remainder stagnant.

Fomzulac f o r towiis
I n a genetic approach to urban geography we are chiefly interested in the
town as an evolving organism. W e want to know its present pattern, how
that developed, and where a i d why the town started. For the present we can
ignore the street plan, the source of its trade, the extent of its uniland, and
a vast nuinber of other interesting features of a city. I see tlie city as a
series of zones, expanding from the original nucleus, and usually conditioned
in its growth by the surrounding environment, unless we are dealing with a
city of the plains. Several features therefore seem to be essential in our
proposed formulae. W e should know something as to the original site or
 60 TAYLOR-GENETIC U R B A N GEOGRAPIIY [ Mar.

nucleus; as to the reason for its founding; as to its growth with respect to
the nucleus. As regards the zones, we should know their width aiiti position.
Further we might well include some clue to its stage of development, and of
ccurse a reference to the total population.
I n 1938 I read a paper before this Association dealing with a number of
towns in the Italian region of the Trentino. There 1 gave a preliminary
description of forinulae, including many of these terms. Since that time I
have been trying to apply the same technique to American atid other towiis.
I feel sure that such research is worth Ti-liile. ant1 equally sure that the
technique which is suggested herewith will be greatly iinproved upon a s
time goes on.
Downloaded by [New York University] at 06:16 15 February 2015

Abbrcvinfioiis ILW/ i i i f l i p forrridac

Directioiis. S o r t h (S.j , Xortlieast (S.E.), etc. ; all from the nucleus
(Xu).
Stage of Dmdopriicrit. Infantile ( I n j , Juvenile (Jii j , ;Idolescent ( A d ) ,
Mature ( M a ) , Senile (Se) , Sucleus ( N u j .
Site of f o z w . Capital ( C p ) , Fan ( F n ) , Fort ( F t ) . AIountain ( h I t ) ,
Mining ( M n ) , Pass (Ps) , Oasis (Oa) , Railway ( K y ) , Rail-crossing
(Rc) . Road (Rd) , River ( R v ) .
Populatiori. Figures represent tliousands.
Distariccs of cories, etc. Figures represent kilometres.
Z o ~ e s . Apartments (Ap), Factories ( F c ) .
Houses-Lai-gest (mansions) (Ha j ; Large (Hb) ; Siiiall ( H c ) ;
Shacks or decayed houses (Hd) .
Offices (Of) ; Shops (Sli) ; Stores, large ( S a ) .
Equation order. Population, Status, Site .. . on tlie left.
Zones, proceeding outwards on the right.
Let us run over the dexription of Port Credit and see how our formula
would work out (Fig. 18). The population is about 2000. This would be
represented by ”2” (the nuinher of thousands) in the formula. The town
was originally a lake-port on a little river. This is indicated by Rv Pt. l17ith
regard to the nitcleits, the river is to the west, arid the lake (or sea) to tlie
southeast. So our formula becomes “Kv(M’.) P t (S.E.) .” T h e stage of
developiuent is hetween atlule?cent and early iiiature. Since it is so sinall a
town, and the mansions are linked to Toronto rather than to Port Credit, it
would seem to be adolescent This general portion of the formula
forms the left side of our equation.
Now we may turn to the particular zones, which appear in the right-lzarttl
side of the equation. The present shopping centre has been shifted from the
original nucleus, about 3 kilometre to the north. Thus the shop-zone appears
as ‘‘Sh(SN.).” The house-zones should he given in order from the centre
of the town outwards. T h i s the second-class houses (Hbj form a zone with
 19421 FORMULAE FOR CLASSIFICATION 61

a centre about one kilometre northeast of the nucleus, i e . , H b ( lN.E.).

Then conies Hc, the original large wooden houses which still surround the
nucleus, but also cover a zone one kilometre to tlie north, i.e., H c ( N u + 1N.).
Tlie shacks H d are fouiid one kiloiiietre west of tlie nucleus, and also 2 kilo-
metres north and so lye get H d (1W. -1- 2N.). Finally the large brick houses
(not quite mansions) represent class Ha, and are being built 2 kilometres
northeast of the nucleus, i.c., Ha(ZN.E.),
The total foriiiula comprises two sides of an equation as follows :
2 A d + K v ( W . ) P t ( S . E . ) =Sli(+N.) + liHb(1N.E.) + H c ( N u i - 1N.) i-
H d ( 1W. ;- 2N.) + Ha(2N.E.)
The factories are not arranged in a zone, hence are not mentioned in the
Downloaded by [New York University] at 06:16 15 February 2015

formula ; hut tlie presence of a few factories may be gathered from tlie fact
that the town is adolescent.
To reconstruct a rough plan from the foriiiula, we proceed a s follows
(see the inset in Fig. 18). A large clot is placed for the nucleus, and con-
centric circles are drawn with radii increasing by half kilometres. W e can
add the river coming in from the west, and tlie lake shore along the south.
The distance indices (in the brackets) show the position of tlie centre of a
zone. T l i ~ i sthe shops can be placed half a kilometre to the north of the
nucleus. If reasoiiable patches or zones be drawn contiguous to each other,
in accord with the indices, we obtain a plan which is quite close to the actual
pattern of Port Credit, aiitl gives us almost all the characteristic features of
the settlement.
The formula for ‘II7hitbycan be obtained in the same fashion. I t conies
out somewhat as follun-s :
4.5 ,4clLCrPt(3S.) =Sli(Xu) +Hb(iE.-c+W.)cHc($S.-c 1;s.) -t
Hd(3N. + 1s. r 2;s.)
l y e may interpret this as follows. A small town in the adolescent stage
with a ppulation of 4.500 Iias develo1)ed at a cross-roads with a little port
3 kilometres to tlie south. O n the right of the equation we find the zones;
i.c., the shops at the nucleus, with three house-zones a t various distances a s
shown in the brackets. F o r instance there are three separate areas of Hd
(siiiallest houses) : at one-half kilometre north, one kiloiiietre south, aiid again
at 24 kilometres south. A reconstruction from tlie formula is given as ail
inset in Figure 19. I t clearly gives us a fair picture of the actual plan of
Whitby.
Forwzulac fov Toronto. Turning now to the four diagraiiis given in
Figure 16 we may reduce the zones to formulae somewhat a s follows :
Toronto was priiiiarily a harbour, and the sandy hook (shown in Fig. 15)
was probably inore vital in 1793 than it would be now. F o r instance, such
important ports as Chicago and Gary have been developed where natural
 62 TAYLOR-GEKETIC URXAN GEOGKAPIIY [Mar.

FIG. 20.-Three towns in the Italian Alps illustrating Eotechnic origins. Formulae
Downloaded by [New York University] at 06:16 15 February 2015

are given in the text. The circles are half a kilometre apart, with the iiucleus as centre.

harbours \\.ere almost non-existent. (Here we 111ay hear the possibilists

cheering ! !) But as time went on tlie political aspects became of great iiiipor-
tance. Hence we niust indicate that Toronto is a Capital (Cp) i n oiir
formula. Obviously each of the four maps of Toronto (at tlifferent periods)
has a different formula; hut the nucleus is always as intlicatcd by the little
circle in the earliest map.
For the town in 1818 the formula is
1 . 2 I n + I P t ( S . ) C p = H t l ( ~ u-Hb(2147.)
)
F o r the town in 1842 we get
20 J u + P t ( S . ) C p = H c ( N u + ~ W . +) Sh(11V.) iHb(2N.1V.t~N.E.)
F o r the town in 1883
120 E r P t (S.) Cp = H d ( Xu - 211’.) Of ( l1V.) I- H c ( N . + 31V.) +
-1

Hb(35. + 45.11’. - 411‘.)f Ha(2S. t SW.)

F o r the present town in 1910, the formula is naturally a long one.
8 9 0 n r a - P t ( S . ) C l , = j H d ( S u - l E . + 3 1 T 7 , ) - - Of(11V.) + F c ( l $ S . E . +
414‘. A 6N.1V.) - H c [ 7 ( 2 S . E . + 35.1V.) 16N.11r.1 + H b ( 4 E . i
4N. - 7S.1T7.) ~- H a ( 3 X . + 6X.1V. i9X.1V.j
In this formula the area corered by tliird-class houses ( H c ) is in two
definite belts. These are suggested by the doublc brackets in the formula.
Obviously other details may he atltled to the forinula. For instance apart-
ment houses form a characteristic zone (for which we may use the abhre\-i-
ation “Apa-20”) fi\-e kilometres to the northwest of the original nucleus.
This could be shown as “Ap (5S.1Y.) .” Since no figure appears outside the
“direction” bracket. 11-e should plot it as a patch less than kilometre wide.
1Ve may use the contraction “Nort-zo” for tlie cliaracteristic house-zones, and
so for the other leading features which detcriiiine the character of the tow1
and formula.
The technique will become more familiar if we consider a set of cluite
 19421 FORMULAE FOR CERTAIN ITALIAN TO\VNS 63

dit-feretit settlements which I investigated in 1938 in Italy. I give simplified

plans of three of these in Figure 20, as well as tlie formulae deduced there-
from. Trento is a town which originated in Roman times. The present
town had as a nucleus the medieval castle ( F t , i.r.. F o r t ) which commanded
the wide glacial trough of the Adige River.’; It was, till lately, a typical
walled (Eotechnic) triwii ; while the modern town is of the adolescent Paleo-
technic type (Photo 19). Its present poptilation is about 60,000. Hence
i i
Downloaded by [New York University] at 06:16 15 February 2015

PHOTO 1Y.-The town of Trento in the .4Ipine portion of Italy. Notice the .Adige
River below tlie great cliffs. and the ancient Castle at the right of the older part of the
town, which is now surrouiided by modern buildings. The fortified bluff with the
Battista Tomb appears at the left. The view is toward the northwest.

the left half of our “frirmula equation” cnnies out

60 ( E o i A d ) + R v ( l \ I ‘ . ) F t ( N u )
T h e zones of dwellings in Trento a r e soiiiewhat complicated 1))- the fact
that tlie industrial tuwn (Ad ) has spread out beyond the medieval town
(Eo). If we draw the half kilometre circles about the nucleus, we siininiarize
the “Hoir-zo” as follows ((in the right o f the equation).
Hb(3S.W.) +2Ap(S.) +Ha(S.E.) +Of(lfS.lT‘.) Hc(I$N.\I‘.)
There is no well-develuped belt of industries, though a f e w factories occur in
tlie northeast. .4 modernist tourist quarter occurs in the heart of the Hc
zone ; but is too mall to be represented in the formula.
1.7 “Trento to the Reschen Pass,” Griffith Taylor. GPO:/.Rcrt., 30 (April, 1940).
 64 TAYLOR--GE N ETI c' U R R A N GEOGRAP H l
' R[ar.

Bolzano may lie taken as tlie second esaniple (Fig. 20). It is situated
near the juiicti~n (Jri) o f tlie Xtlige and the Isarco (which rises 111 tlie
Breiirier Pass). I t is Iiuilt 011 the large cletrital fan of tlie Tallera Torrent.
Like Tretitii it is half Eotechnic and halt' I'aleotechiiic in pattern. The
"Him-zo" a r e p e r h a p more regular in UIil~aiic),a i d the foriiida is as
fl)IlO\V> :

40 ( E ~ - . l d )-FIIJII(I;S.\\..) = H c ( S u ) Of(+S.W.) i
A~ll)(+\\'.- i N . E . ) H a ( 111'. - lN.E.)
The little village of Lai\es \I it11 a ~ ) o l ) t i l a t i o i iof al)out 500 has grcnvii up
on a fan ahlllt 8 inlles smith of Bolzanu (Fig. 20). It is less than one
kilometre across (Photo 20). Here u e f i n t l merely a series of aliout 60
Downloaded by [New York University] at 06:16 15 February 2015

i i

PHOTO 20.-Laivrs, a small nritcli-village Imilt oii an alluvial fan in the -4dige Cor-
ridor ahout X miles south of Bolzanu. The viiieyards af 1.aives apliear in the fcirr-
ground. The view is lotiking south.

houses along tlie inain ri iail, niostly large and well-built and standing in their
owii vineyartls. There is no tlifferentiatioii into zoiic's. so that it i s in the
infantile stage. ant1 the Iorinula is simple. F(or Laives :
O..; I n F n = HI)(SN.E. + iS.\T'.)
Enough exaiiiples have lxen given for the reader to realise that it is quite
pi.)ssible to suiniiiarise in a one-line forinula tlie salient features iuf a tmvn.
The general cliaracters ( i.r.. site and status) appear on the left of the
foriiiula equation, the zi)iiaI details on the right. I have not atteiiipted t o
 19421 POSSIBILISM I N RACE AND NATION 65

cover any but paleotechnic towns (ie., the usual occidental industrial towns) ;
but of those referred to earlier in this address, Pekin is oriental with some
senile characters ; while Biskra is partly paleotechnic and partly oriental.
Urandanji, McMurray, and Waterways are infantile. Canberra with its
cobweb patterns and its isolated communities is neotechnic, approaching the
biotechnic town ; but lack of space forbids my discussing any but the paleo-
technic type of town.
SECTION G
POSSIBILISM APPLIED TO RACE, NATION AND CITY

As most of my readers know, I have always been a rather definite environ-

mentalist. In concluding this address, I wish to consider whether deter-
Downloaded by [New York University] at 06:16 15 February 2015

minism or possibilism is of more importance in connection with the three

types of human groups which I have studied with some thoroughness. I n
my book on the major human races (Environment,Race and Migration, cit.),
I advanced data which seemed to me to prove that in the remote past races
had actually differentiated as the result of environmental control and environ-
mental changes. I n those remote periods (certainly prior to 20,000 years
ago) man migrated primarily for the same reasons as the higher mammals.
H e exercised little choice, and a geographer in those days would almost
certainly have been a whole-hearted determinist ! Thus the major human
clusters in the far past were certainly chiefly controlled by environment.
I n the later historic period, the major migrations of earlier times, and the
mass-migrations of today, seem to me to illustrate the push from the home-
land combined with the pull of the pioneer lands. If a man experiences
neither push nor pull, he stays at home. A study of the environmental con-
trols of various parts of the world, and the way in which these may affect
migrations of the future was published by the writer in 1922.16
Turning now to the smaller groups such as the nations of Europe, I
studied the reasons for the distributions of cultures, towns and populations
in that continent. It seemed to me clear than those anthropomorphic ser-
vants of the environment, King Coal, King Frost, and King Drought, were
far more important in producing the population-pattern of saturated Europe
than were the actions and orders of men, from Charlemagne and Napoleon
downward. The widespread changes due to human orders (ie., “choice”)
produced permanent results only if they were in line with environmental con-
trol, but if not, such changes were soon dissipated or annulled. (Environ-
ment and Nation, cit.) .
In a general sense this is true of major cultures as of major populations.
Roumanian culture has persisted chiefly owing to the ever-present refuge
l6 “Distribution of Future White Settlement” Geog. Rev., 12 (July, 1922).
 66 TAYLOR-ENETIC URBAN GEOGRAPHY [Mar.

given by the southern Carpathian plateaux. Italian culture is the creation

of the northern Alpine barrier. Swiss culture clings to its Alpine refuge.
French culture is Roman culture, bounded by the Rhine barrier and the
Rhine defence. German culture is in large part the non-Roman culture
preserved in the thick forests of the west German plain. British culture is
due to her sea-girt island.
Yet it is clear to the impartial student that as the human clusters become
smaller, the general effects of environmental control on the unit become
intermingled with rather special results due to man himself. Thousands of
worthy Americans live in Indiana today, partly because (as the possibilists
say) the ancestral emigrants “chose” to leave England. But what seems an
Downloaded by [New York University] at 06:16 15 February 2015

even more potent reason to a determinist is that the environment favoured

large and prosperous families. Other quite similar emigrants, in the distant
past, chose to migrate to the West Indies, but their numbers have remained
negligible ; for environmental reasons as I see it.
Ratzel was the great German protagonist of determinism ; and I have
often wondered if the growth of Possibilisni in France was not in some
measure due to a national antipathy to all things German. Indeed I have
seen the statement that Haushofer’s modern Geopolitics (extreme Imperial-
ism) is a logical outcome of Ratzel’s Determinism. Could not one as well
derive the many dangerous schisms in French politics from the individualism
of French Possibilisuut! I illustrate the effect of environment upon one’s
philosophic creed by an exaggerated reference to certain American geog-
raphers.
I ask you to pity our unfortunate colleagues who have been brought up in
that part of the Eastern States whimsically called the “Middle West.” The
region between Chicago, Omaha and St. Louis has so uniform a climate and
so level a topography that major environmental controls are necessarily
almost lacking. Moreover the general conditions were so favourable to
settlement (at any rate in the early days) that wise, medium, or foolish,
choices of settlement-technicques all resulted in a fair measure of prosperity.
How could our worthy Middle West colleagues escape becoming Possibilists !
I said I was exaggerating, but I do firmly believe that on some nine-tenths
of the earth’s surface man’s choice is so limited as to be almost negligible.
There is usually only one way to exploit the terrain most successfully, and it
can only be ascertained by lengthy and thorough investigation of the environ-
ment.
The third of our categories concerns town and cities. Here we are deal-
ing with man-made structures, and it is evident that man can choose every
detail of the object of his own creating. Here I grant that possibilism plays
a role of greater importance that does the environment. Yet even in the
 19421 POSSIBILISM IN CITY A N D TOWN 67
innumerable towns of the fertile plains of the Middle West-where the chief
feature of a suburb is said to depend on which side of the railway it lies-even
in these checkerboard towns there are many problems which depend very
largely on the environment.
I often wonder why there is so much opposition by geographers to the
concept of environmental control. One would think little of a doctor who
spent ten years studying medicine, and then practised according to the tenets
of Mary Baker Eddy. Or of a lawyer who summed up his studies in the
words of Bumble the Beadle, and loudly proclaimed that the “Law is an Ass.”
A student of the widespread sterile areas in the world inevitably tends
to become somewhat of a determinist. It is precisely because the student
Downloaded by [New York University] at 06:16 15 February 2015

of Urban Geography is engaged in the field which is least affected by environ-

mental control that I have dwelt on possibilism in this address. Man, like
every other living creature, is a parasite on the surface of the earth. He
exists by exploiting the environment. For me, the answer to our con-
troversy (possibilism versus determinism) is suggested by a modification of
the old proverb, “Nature pays the piper, and, in most parts of the world, she
also calls the tune.”
A dozen years ago I was in Shanghai, and made some study of Sun Yat
Sen’s book “International Development of ,China.” I was much impressed
to find that this book, by the leader of the young Republican Party of
China, was largely a study of the economic geography of that country.
W e see the same thing in the plans of the Soviet Union. Some day our
discipline will progress beyond its present stage of the Cinderella of the
sciences. Cannot we geographers study the world to such purpose that we
shall in time be accepted as a body well suited to guide the nation along the
paths of material and cultural progress? Can we not all strive to become
Nation-Planners ?17
May the day come when the public will say of every careful student of
our discipline, “He has studied material progress all over the world. H e
knows the possibilities, and can interpret our environment to our best benefit.
I n a word he is a competent geographer. We must listen to him.”
Acknowledgements. I have to thank a number of people for supplying
me with data for this address. Among these are Kenneth Binns (Canberra),
G. Gordon (Port Credit), Nadine Hooper (Toronto), A. Pollan (Boston),
F. T. Rowe (Whitby), and George Tatham (Toronto).
University of Tormto
December, 1941.
1 7 “The Geographer’s Aid in Nation-Planning,” Griffith Taylor, Scot. Geog. Mag.,

Jan.-March, 1932.