An Experimental Study of Somatic Modifications etc. 
321 
ent parentage. Assuming a given modification of the value and 
supposing that ^ represents the proportional part of this to be trans- 
mitted, the offspring of the two lots would be found to differ by the 
X' 
quantity ^. If the second of these methods were employed, we 
should compare the second generation with the parent generation, the 
two being measured at the same age. Assuming as before that the 
parents had been modified to the extent cr, the offspring, according 
X 
to hypothesis, would be found to be modified to the extent + — 5 
i. e., the effect of the conditions would have been cumulative. 
Now, as a matter of fact, I have attempted both of these tests 
upon rather a large scale; but I have not yet found the second one 
to be practicable, owing to the difficulty, without special facilities, 
of repeating precisely the same temperature conditions during the 
lives of two successive generations But, the first of these tests 
finally proved to be practicable, and has yielded the results which 
are dealt with on the ensuing pages. 
The methods which I have employed in dealing with the parent 
mice have been described in detail in my paper of 1909, so that 
little attention need be devoted to them at present. The parent 
generation (i. e. that immediately subjected to the experimental 
conditions) which is here to be considered is the one which I have 
discussed on pp. 139—144 of the paper cited, under the heading 
» Series of 1908 — 1909*. At the date of writing that account, however, 
these mice had not yet reached maturity. The temperature conditions 
to which they were subjected from the commencement of life 2) until 
some days before the birth of their last brood of offspring are represented 
in Fig. 1 of the present paper. The mean temperatures of the warm 
and cold rooms respectively throughout this entire period were 21.27° 
and 5.57° C. respectively. For the period prior to the date of pairing 
(May 2), the mean difference was somewhat greater than this, viz. 
16.8° C. Since the curves shown in the chart are based upon weekly 
averages, the extremes of temperature are of course eliminated. The 
1) See foot-note on p. 324 and 325 below. 
2) Sometimes commencing with the day of birth, sometimes a few days 
after this, in many cases before birth i. e. during the pregnancy of the mother. 
3) I am indebted to my wife for the tedious work of compiling these data 
from the thermograph sheets. 
Arcbiv f. Entwickluagsmeclianik. XXX. 2. 21 
