THE CAUSES OF VARIATION. 233 



perature at the time of impregnation that a difference of a 

 few degrees of warmth or cold lasting for only a minute at 

 this time has as o-reat an influence as a similar difference 

 extending over the whole eight days of subsequent develop- 

 ment. To come to figures, it was found that as a mean of 

 eight sets of observations, ova impregnated at about 8° C. 

 were 4*2 per cent, smaller than those impregnated at from 

 1 7 '6° to 2 1 '6°, the subsequent conditions of development 

 being identical. Also, as a mean of five observations, 

 those impregnated at about 25° C. were 5*2 per cent, smaller. 

 In all these observations the ova were kept in the cold or 

 w^armed water for an hour at the time of impregnation, but 

 in seven other experiments they were kept for only a 

 minute at 8° or 25°, and were then transferred to water at a 

 normal temperature. Nevertheless the larvae were, after 

 eight days' growth, on an average 4"i per cent, smaller. 

 In four instances the ova were kept for only ten seconds 

 during impregnation at the abnormal temperature, the 

 larvae being- then on an average only i 7 per cent, smaller. 

 We may presume that the time was so short that less than 

 half the ova were in these cases impregnated at the abnormal 

 temperature. 



In addition to the observations cited, about thirty other 

 sets of determinations on the effect of temperature at the 

 time of impregnation were made, some of them at tem- 

 peratures, as 15° and 23*5°, intermediate between the 

 extremes mentioned above, and others of them in 

 which the larvae were measured after four, five or six 

 days' development. They all serve to confirm the fact of 

 the special sensitiveness of the ova to temperature at the 

 time of impregnation. 



Experiments were also made on the effect of temperature 

 during the course of development. Larvae kept at 23° to 

 24° for the whole period of eight days were only about 2 

 per cent, smaller than those kept at 20° ; were, in fact, affected 

 to just about the same extent as when the temperature of 

 impregnation was 23° to 24°. Also larvae kept at about 

 15° during development were about 4 per cent, smaller than 

 others kept at 19°. 



