248 StitfJirs ill the PhjiHioIiKin of Fcrtilizatioii 



eggs fertilized after ;i previous stay of four lioiirs in srawater hatched 

 quickest of all, just as the 4-cell division of these eggs had been com- 

 pleted in the shortest period. 



In comparing the last column of Table X, whieJi givi's the condition 

 of the larvae, with that in which the times of hatching are recorded, it 

 is seen that those larvae which emerge from the c^gg in the shortest 

 time from fertilization are the healthiest. 



In the four experiments of Series B the i-ates of early segmentation 

 were noted, but not the times of hatching. The general result is th(' 

 same as that of Series A, namely that the later fertilized eggs segment 

 quicker, but Exp. 8 shows a marked irregularity difficult to account for. 

 There is a fall from the 91 minutes of the 1^-hour eggs to 80 minutes 

 of the 2|-hour eggs, after which the next fertilized lot give 87 minutes, 

 and then the final batch drop to 76 minutes again. Finally in Exp. 6, 

 there is a decrease in the times at which the 4-cell stage was attained 

 from the 1-hour lot (99 minutes) to the 2^-hour lot (81 minutes), after 

 which there is a rise to 90 minutes again. This is a phenomenon 

 similar to the change in the hatching rates in Series A. The segmen- 

 tation time of the last fertilization in Exp. 5 is given in brackets, as it 

 is not really justifiable to compare this observation, made on the very 

 few eggs that were fertilized in this case, with the large numbers in the 

 other batches. 



Exps. 5 and 7 show that the increasi' of segmentation rates with 

 succeeding fertilizations went hand in hand with a decrease in the per- 

 centages of eggs fertilized. 



The experiments of Series C show an exactly reverse phenomenon 

 to those of Series A and B. In every case the segmentation rates 

 become slinver with successive fertilizations. This is especially marked 

 in Exp. 9, where there was a gi-adual rise from 82 minutes of the 1-hour 

 fertilization to 93 minutes of the 5-hour lot. The two experiments 

 forming Series D behaved similarly, although there was a final increase 

 in the rate for the last fertilization of Exp). 13. 



It was thought that there might be a connection between the 

 change in temperature during the progress of the experiments and the 

 alterations in the rate of segmentation. The evidence, however, seems 

 to point rather against this. In Series A the temj)erature rose from 

 19*5" to 21-5° C. during the experiments. In Series C it remained 

 constant, and in Series D rose from 18 to 18'5"C. If the rise of two 

 degrees in the experiments of Series A will account for the increase in 

 the segmentation rates, there is no fall in temperature in the later 



