60 X. KTSPIIXOUYE. 



are uncertain I .shall call them tlie lafevid humps. The second lateral 

 longitudinal grooves had now reached the posterior end of the fifth 

 appendage- hearing segment. The seventli appendage-bearing segment 

 had been newly cut off. At this time all tlie appendages are post-oral. 



The external changes that occur before the appearance of the 

 rudiments of all tlie ambulatory appendages are very hard to observe, 

 as the \'entral plate is small and pale at this time. I could not observe 

 them in the summer of 1889, except the irregitlar fissures found in a 

 very early stage. 



As the result of the gradual enlarging of the veritral plate, the 

 ego^ gradnall}^ flattens on the ventral side, thus producing a space 

 filled with a clear fluid between the ventral ])late and the vicarious 

 chorion whicli is by this time already present. The vicarious chorion 

 dilates greatly while tlie clKirion outside it cannot expand ecjually ; 

 thus the chorion is necessarily ruptured. T^wo specimens of the 

 Ijroken chorion are represented in fig. 11. The lines of rupture are 

 seen to o'O round the chorion irreo'ularlv and do not form closed 

 circles. 



On August 15tli (18 days after fertilization), the eightli appendage- 

 bearing segment had appeared. The second lateral longitudinal 

 groove had reached the posterior end of the sixth appendage-beaTing 

 seo-ment, bevond which the o-voove never elono-ates. The seventh 

 appendage had appeared on this dav. 



On August 16th (19 days after fertilization), the nintii append- 

 age-bearing segment had appeared (fig. 12). The eighth appendage 

 had been produced. The median longitudinal groove, whicli had 

 appeared four days before in the posterior part of the ventral plate, had 

 become very distinct in the List undifferentiated segment. The lateral 

 humps were not as elevated as before. 



On August 17th (20 days after fertilization), the tenth appendage- 



