loG Sjimmetry of Egg and Symmetry of Embryo in the Frog 
In the positive series tlie range of 7 diminishes from I. to IV. ; and the mean 
value of a = /3 when 7 = 0 diminishes too. The rate of change of the mean value 
of yS is greater than that of a as 7 alters in I., equal to it in II., less than it in III. 
and still less in IV. 
In the positive series the moan value of /S is of course less than that of a only 
while 7 is positive. In the negative series it is less than that of a as long 
as 7 does not exceed a certain small value in II. III. and IV. : in I. it will be 
greater than that of a as long as 7 does not fall below 55°. 
Lastly, in Table XIII. are given the variabilities of each angle when one other 
is constant. 
TABLE XIIL 
Positive Series 
Negative Series 
.Vaiiabilities 
of a or /3 
wlien 7 is 
constant 
of /3 or 7 
when a is 
constant 
of a or 7 
when /? is 
constant 
of a or ^ 
when 7 is 
constant 
of /3 or 7 
when a is 
constant 
of a or 7 
when ^ is 
coustant 
I. 
Eggs 
close. Axes horizontal 
1-499 
2-140 
1-483 
1-305 
2-579 
1-698 
II. 
Eggs 
close. Axes vertical . . . 
1 -308 
1 -490 
1-489 
1-224 
1 -863 
2-226 
III. 
Eggs 
spaced. Axes horizontal 
1-179 
1-254 
1-443 
0-746 
1-308 
1-553 
IV. 
Eggs 
spaced. Axes vertical 
1-057 
1-066 
1 -.337 
0-976 
1-344 
1-494 
When the eggs are close and their axes horizontal (I.) it is evident, from both 
the positive and negative series, that the variability of /3 when a is constant is 
greater than that of a when 13 is constant. 
When the influence of gravity is removed (II.) the variability of /3, a being . 
constant, is equal to that of a, 13 being constant, in the positive series, but in the 
negative series that of is less than that of a. Under other circumstances (III. and 
IV.) and in both positive and negative series, /3 is less variable than o (a or /8 
being constant). 
Again, comparing ^ and 7 in the same way, j3 is slightly more variable than 7 
in I. (positive series) but less variable in all other cases. Lastly, a is less-variable 
than 7. 
It must not be forgotten, however, that in this method 7 will have a much wider 
range of variability than in that first employed by me, in which the range was 
restricted to 90°. The range of a is, however, limited by our convention to 90°. 
Comparisons in respect of variability may therefore be more fairly instituted 
between /3 and 7 than between a and either /3 or 7. 
It will also be noticed that the variability diminishes as the interfering agents 
are removed. 
