BLASTOGENIC VARIATIONS. 



109 



artificial fertilisations were carried out at short inter- 

 vals over periods of several months, and the larvse al- 

 lowed to develop under conditions which were probably 

 nearly constant, except as regards temperature. For 

 this varying factor a correction could be easily applied. 

 In spite of the constancy of environmental conditions, 

 however, the size of the larvss showed very marked 



Apr. Apr. M y June July July Aug. Sept. Oct. Nov. Mov. Dec. Jan. 

 Ifet 35th 19th 12th 6th .80th, 23cd 16th JOth Scd jWih ;30th 13th 



FIG. 19. Seasonal variation in size of larvae. 



variations. The range of these variations may be 

 gathered from the accompanying diagram. Here the 

 ordinates represent the mean body lengths of the larvse 

 in micrometer eyepiece units (of which 152.3 are 

 equivalent to 1 mm.), and the abscissas the time of year 

 at which the fertilisations were made. It will be seen 

 that in April and May the larvae were on an average 

 about 34 units in length, but that then they began 

 steadily to dwindle down in size, so that in June they 

 were about 31 units, and in July and August only about 



