POULTRY SCIENCE

July, 1940, Vol. XIX, No. 4

Percentage Shell as a Function of Shell

Thickness, Egg Volume, and
Egg Shape
V. S. ASMUNDSON AND G. A. BAKER
Divisions of Poultry Husbandry and of Mathematics and Physics, University of California, Davis
(Received for publication October 20, 1939)
/ V
"T~ HE avian egg may be divided into inherited difference in the thickness of
J- four parts: the yolk, the albumen or the shell. They concluded that percentage
white, the shell membrane, and the shell. of shell was, for their purposes, a satis-
The absolute and relative weight of the factory measure of amount of shell when
yolk, the albumen, and the shell varies compared with measures requiring more
for different species and also within nar- elaborate computation such as that of
rower limits for eggs of the same species Hendricks, Lee, and Godfrey (1931). Per-
as shown by the summary of Grossfeld centage of shell may not be a satisfactory
(1938, pp. 70-78). Asmundson (1939) has measure of amount of shell when compar-
also shown that the absolute and relative ing the eggs of different species if the
weight of the shell membrane varies for thickness varies significantly. To determine
different species, the turkey (Meleagris whether there are species differences in
gallopavo) egg having a heavier shell mem- the thickness of the shell the data in
brane than the chicken (Gallus demesti- Table 1 were compiled. From the data
cus) egg. There was also some indication given in Table 1 it is apparent that the
that the shell on the larger turkey eggs shell of the ringneck pheasant egg is
may weigh relatively less than the shell on thinner than that of the turkey egg but
the smaller chicken eggs, although the shell the percentage of shell on the former
on turkey eggs is thicker, on the average, is nevertheless greater than on the latter.
than the shell on the chicken eggs. This Other groups of turkey eggs, however, had
finding prompted us to inquire into the up to 9.26 percent of shell on the average,
influence of egg volume, egg shape, and a value which is still slightly lower than
shell thickness on percentage of shell. for the chickens and pheasants. The thick-
ness of the chicken egg shells in Table 1
SPECIES DIFFERENCES IN THE is not known, but Hays and Sumbardo
THICKNESS OF T H E SHELL (1927) have shown that the thickness of
Taylor and Lerner (1939), using White the shell varies from .280 to .345 mm.
Leghorn chickens, have demonstrated an on eggs from different hens, or an average
[227]
228 V. S. ASMUNDSON AND G. A. BAKER

of .309 mm. An analysis of variance of EFFECT OF EGG VOLUME AND EGG

the original data for ringneck pheasant SHAPE ON PERCENTAGE OF SHELL
(Phasianus colchiens torquatus) and tur- Pearl and Surface (1914) have found
keys summarized in Table 1 and the data that an egg can be represented as a perfect
published by Hays and Sumbardo (1927) prolate spheroid, as far as volume is con-
shows that the thickness of the shell dif- cerned, with only an error of 2.2 percent
fers significantly for these three species. on the average. This error is an excess.
The few silver pheasants (Gennaeus nyc- We shall be concerned with the ratio of
themerus) eggs available had shells of surface area of an egg to its volume. If
about the same thickness as those on tur- the volume is slightly in excess the surface
key eggs. area will probably be in excess also, so

TABLE 1.—Data for the eggs of jour species

Shell
No. Weight Length-
of of breadth Weight Thickness
eggs egg index
egg shell grams percent mm.
Ringneck pheasant 23 26.66 78.68 1.085 2.325 2.58 9.68 0.26
Silver pheasant 7 44.09 75.07 1.083 2.288 4.88 11.07 0.36
Chickens 64* 55.81 72.68 1.082 2.235 5.28 9.46
Turkeys 64 90.37 74.01 1.075 2.174 7.48 8.28 0.35

* Average for eggs from each of 64 hens or 384 eggs.

The specific gravity was determined by that the ratio of these quantities calcu-
obtaining the weight W in air and the lated for a prolate spheroid should give a
weight Ww in water. The specific gravity very good result.
was calculated from the formula Let
W P = percentage shell
Specific gravity=
W-W w D s = density of shell
D e = density of egg
The data obtained are summarized in
T s = thickness of shell in centimeters
Table 1. Those for chickens do not dif-
S = surface area of the egg
fer significantly from those obtained by
V = volume of the egg in cubic centi-
Hays and Sumbardo (1927), Olsson
meters
(1934), and others.
Then by definition
The values for the specific gravity of
turkey eggs and shells are slightly lower D . ST,
(1) p=—rxioo.
than those for the other three species but
the significance of the differences is doubt- STS = volume of the egg shell if the shell
ful. Since the thickness of the shells and thickness is T s over all the egg. Now if
the specific gravity of the eggs and shells
will not satisfactorily explain the apparent DsT s is constant then P will depend
differences in the percentage of shell on
the eggs of the four species considered, only on S/V or perhaps this dependence
other factors must be at least partly re- can be stated in terms of V only.
sponsible. A prolate spheroid is generated by re-
 PERCENTAGE SHELL 229

volving the upper half of the ellipse,

(2) V—=l,a>b,
a2 b2
about the x-axis. In (2) a represents one-
half the length and b one-half the width
of the egg. The volume is given by the
equation

(3) V=—rb 2 a.
3

The surface area is given by the equation

4xb ra ,
(4) S = V a « - ( a » - b * ) \ ! dx
a2 Jo
2-ira2b
2
= 2irb -|—-. 2 2 sin
Va -b Fr=-)
If the eggs of different sizes are similar
in a geometric sense, that they are is in-
dicated by figure l, then for all such simi-
lar eggs b = Xa, X < l. FIG. 1. Eggs selected as reasonably typical of
If b = Xa we obtain those laid by birds of four species. They should
not be used to furnish an absolute figure for the
/ sm-Vl-X2\ relative length and breadth of eggs laid by these
species.
(5)
' 2Xa
But tionality being p, then the value of the ratio
V = 4/3 7rb2a = 4/3 7rA2a3 surface area
for the new egg is (1/p) ' k.
so that volume
In (7) P is considered as function of
\4*X2 /
V; that is, X is considered constant. X is a
Hence (S) becomes parameter that determines the shape of
the egg. It may be of interest to regard
/ sin-'Vl— X 2 \
s V Vl-x 2 / P as a function of the shape or X. V will
(•fii
W 2i/3 \ « i v 1 ' 3 then be regarded as constant.
If X is constant (7) is of the form
and (1) becomes
K
(9)
sin-Vl-A2

(7) P= -
yi *"' rx+
)„, Ta 100
V1'3
yvs
where K is constant. If V is constant then
2" 3 V s D„ (7) is of the form
The following proposition can be stated
about eggs, considered as solids of revolu- (10) / survi-x2 \
tion, whether the longitudal cross section
is an ellipse or not. If the ratio, S/V = k, Typical graphs of (9) and (10) are shown
has been found for any shaped egg and in Figures 2 and 3. Figure 2 shows what
the value of this ratio is desired for any happens to the percentage shell as the size
similar shaped egg, the factor of propor- of eggs similar to a typical hen egg,
 230 V. S. ASMUNDSON AND G. A. BAKER

X — j4, changes. Percentage shell changes within ordinary limits. The length-breadth
fairly rapidly with change in volume es- index, calculated for the eggs of SO species
pecially for the smaller eggs. of wild birds (Grossfeld, 1938, pp. 2-3),
Obviously species that lay small eggs ranged from 62.9 percent to 83.8 percent.
must lay eggs with thinner shells than Such variation in shape might cause a vari-
species that lay large eggs or their eggs ation of 1 percent in the shell whereas the
would have a high percentage of shell. actual range is from 3.9 to 12.1, or a dif-
Thus the data presented by Grossfeld ference of 8.2 percent.

/<?

/6

14

> /a
^s
^<0
fc /o ^J^= .035" cm.
$
J.^> s
Vl
^
* 6
rs~.t 726 c,77

O /O £0 30 40 JO 60 70 SO SO /OO //O /20 /30

Vo/c//f?e //? o/6/c cent/meters
FIG. 2. Functional relationship between percentage shell and egg volume for two shell thicknesses,
.035 and .026 centimeters. It is assumed for purposes of illustration that "k = 3/4, T s = .035 cm. or .026
cm., D e = 1.08, D s = 2.26. If these quantities are different the corresponding curve may be ob-
tained from either curve by multiplying by a proper factor. Either curve above may be derived from
the other by multiplying by a factor. If X is quite different say X < 0.5 allowance may have to be made
for this fact.

(1938) indicate that birds laying small Figures 2 and 3 use the values X = %,
eggs; for example, canaries, must lay eggs D e — 1.08, D s = 2.26 for Ringneck and
with relatively thin shells since the per- Silver pheasant, chicken, and turkey eggs.
centage of shell is less than found on the These figures are averages from data in
eggs of gallinaceous birds. Table 1. Figure 2 is based on a shell thick-
Figure 3 shows what happens to per- ness of .035 cm. which applies approxi-
centage shell as the shape of a hen egg mately to turkey eggs and .026 cm. which
of typical volume changes. Percentage shell is the approximate thickness of the shell
is very little affected by changing shape on Ringneck pheasant eggs, Figure 3 uses
 PERCENTAGE SHELL 231

51.7 ex. as the volume of an average If D s = D' s and D e = D' e then it is only
hen egg. necessary to multiply the ordinates of
The equations of the curves in Figure 2 (1) by T' s /T s . The graph of (12) can
are be obtained by multiplying the ordinates
35.92 of (11) by .026/.035.
(11) P=
From Table 1 and Figures 2 and 3 it is
and
26.68
(12) P=-
V'/3
The equation for Figure 3 is
sin-Vl-*2
(13) P = 4.754(X2'3+
X2 /
If D e = 1.08 for each species, then
from Table 1 the average volumes are >
•§ *
•**

24.7, 40.8, 51.7, and 83.7 c.c. The calcu-

lated percentages of shell are 9.2, 10.4,
9.6, and 8.2. The corresponding observed *?
averages from Table 1 are 9.7, 11.1, 9.5,
and 8.3. The agreement is fairly close.
Special values of X, T s , D e , D s , are as-
sumed in (11) and (12). Other values of
these quantities may be appropriate to the
discussion of other eggs. The curves of
Figure 2 corresponding to a new set of A • ratio t>f m'dM of egg to /e/;gt/i ofegg
specified conditions can be obtained by FIG. 3. Functional relationship between per-
multiplying the ordinates of one curve of centage shell and egg shape for a typical chicken
Figure 2, say (11), by a proper factor. egg. It is assumed that T s = .035 cm., D e = 1.08,
Let K of equation (9) be written as Ds = 2.26, and V = SI.7. If other values are
taken the ordinates of this curve must be mul-
tiplied by a suitable factor.
\ Vl-X2 / D,T, apparent that shell thickness and egg vol-
(14) K = 32'V«- 100 •
2>'3X« De ume are by far the more important factors
The part of K involving X changes so influencing the percentage shell. The ques-
slowly and eggs of different species are so tion arises as to which of these has the
similar that for practical purposes it may greater effect when the other factors, egg
be considered constant. density, shell density, and egg shape, are
This leaves considered constant, which they practi-
D8T8 , cally are for eggs of one species.
=h
Let
D„
AP = change in percentage shell
which can be calculated for (11). ATS = change in shell thickness
Suppose new values D' s , T' s and D' e so AV = change in egg volume
that
- = h'. Then
D'.T'.
D' e
To obtain the ordinates corresponding to (15) ^,pil_i^l
h' multiply the ordinates of (11) by h'/h. P L T. 3 VJ
approximately.
232 V. S. ASMUNDSON AND G. A. BAKER

Equation (IS) shows that the effect of These relationships are shown graphically
a relative change in shell thickness is three in Figures 2 and 3.
times as great and opposite in direction Equation (15) shows that within a
to the effect of the same relative change species a relative change in shell thick-
in volume on the relative change in per- ness has three times as great an effect
centage shell. Equation (IS) is a good ap- in the opposite direction as the same rela-
proximation so long as ATS and AV are not tive change in volume on the relative
too large. Thus the approximation should change in percentage shell.
be good within a species.
LITERATURE CITED

SUMMARY Asmundson, V. S., 1939. The formation of the

egg in the oviduct of the turkey (Meleagris
Ringneck pheasant eggs were found to gallopavo). Jour. Exp. Zool. 82:287-304.
have thinner shells than chicken eggs (data Grossfeld, J., 1938. Handbuch der Eierkunde.
of Hays and Sumbardo, 1927) which in Julius Springer, Berlin.
turn had thinner shells than turkey eggs. Hays, F. A., and A. A. Sumbardo, 1927. Physical
characters of eggs in relation to hatchability.
The differences were statistically signifi-
Poul. Sci. 6:196-200.
cant. The few Silver pheasant eggs exam- Hendricks, W. A., A. R. Lee, and A. B. Godfrey,
ined had shells of the same thickness as 1931. Effects of cod liver oil and ultra-violet
turkey eggs. irradiation, as influenced by oyster shell in the
Calculations based on data for eggs of diet of confined laying hens. Jour. Agri. Res.
43 :517-S3S.
gallinaceous birds show that when the
Olsson, N., 1934. Studies on specific gravity of
thickness of the shell remains constant, hens' eggs. Otto Harrassowitz, Leipzig.
(a) the shape of the egg has a negligible Pearl, R., and F. M. Surface, 1914. A biometrical
effect on the percentage of shell which, study of egg production in the domestic fowl.
for eggs of average shape, decreases slight- III. Variation and correlation in the physical
ly with an increase in the relative breadth characters of the eggs. U.S.D.A. Bur. Anim.
Ind. Bull. 110, pt. 3.
of the egg, but (b) the percentage of
Taylor, L. W., and I. M. Lerner, 1939. Inheri-
shell decreases significantly with an in- tance of eggshell thickness in White Leghorn
crease in the volume or size of the egg. pullets. Jour. Agri. Res. 58:383-396.