1910] Girault, Periods of Embryonic Development. 19 



for the whole deposit than for the maximum period of the lots 

 taken separately. Fourthly, the period of hatching is much 

 greater than the period of deposition, when the former is relatively 

 short at least. 



Coming to the period of embryonic development. First in 

 regard to what was found in the cases cited. Referring again to 

 table I, it is found that variation occurs under conditions where 

 it is ordinarily unlooked for, namely in depositions of nearly 

 simultaneous occurrence kept in similar environments ; but in the 

 nine cases did not exceed a half day by very many hours ; it did 

 not differ from the average by more than slightly over a third 

 of a day nor the mean by more than 9.6 hours. Moreover, it was 

 found that the variation within six hours as between the limits of 

 that period was nearly equal to the variation as between two of 

 the lots deposited slightly over an hour apart (lots VIII and IX) ; 

 whereas in two lots deposited nearly an hour apart, there was no 

 variation (lots II and IV) ; in two lots deposited 11 minutes apart 

 there was practically none (lots VI and VII), but in two lots 

 deposited 4 minutes apart there was a half-day's variation (lots 

 VII and VIII). 



How should the period of embryonic development be 

 measured? Obviously, but in one way and that very simple. It 

 is evident that if measured with but a single case, the period has 

 not been obtained as accurately as if measurement was based on 

 many cases in the same lot because of the random variation of the 

 period of hatching. It is also evident, because of the latter, that 

 the more cases upon which the measurement of any single period 

 of hatching is based, the more accurate will be the determination 

 of the period of development. The question is, then, simply one 

 of determining the best measure of the time involved. The time 

 at which the modes of hatching occur in cases such as this are 

 not very precise, because the true mode will be seldom obtained, 

 and because the modal time does not take into consideration 

 extreme or even intermediate hatchings. Under the circum- 

 stances, the mean time is better and more easily obtained and will 

 be precise enough for all purposes; the average time could be 

 obtained very difficultly only. 



In order to determine accurately the period of embryonic 

 development it is merely necessary to know the limits of the 



