EARLY CHAXGES IX THE OlTJf. 239 



vesicle, exactly as they are figured by Graber [114] (see PL XIII., 

 Fig. 2). 



I conclude that the earlier cleavage of the polar cells occurs 

 in the Blow-fly during the passage of the germ through the 

 food-yelk. If a young morula is formed as I suggest, it is easy 

 to understand the difficulties which arise in the direct observa- 

 tion of the cleavage of the germ-yelk ; and I believe that the 

 morula is easily broken up into its constituent cellular elements, 

 which become scattered in the abundant food-yelk. This view 

 accounts for the segmenting nuclear spindles which have been 

 detected by the laborious researches of Blochmann [107] , and 

 others, lying scattered in the food-yelk. If the young morula 

 assumes the form of an open cylinder, as it does in Synapta, or 

 is perforated by intercellular openings, as it is in Echinus 

 (Selenka [101]), or remains an open cup, it is not difficult to 

 understand the manner in which the food-yelk passes into its 

 interior. 



The view that such is the origin of the blastoderm vesicle is 

 directly supported by Graber's figures [114, PI. VI., Figs. 60 to 

 67], and by his woodcut on p. 260, which I have reproduced in 

 PI. XII., Fig. 12. It is true that Graber regards these figures 

 as representing abnormal phenomena, but I regard them as 

 normal, and am at a loss to understand why he considers 

 them abnormal ; his only ground for the supposition appears 

 to be that they do not accord well with received views. 



Moreover, Metschnikoffs [98] description of the formation 

 of the blastoderm in the Scorpion exactly corresponds, so far as 

 the segmentation phenomena are concerned, with the descrip- 

 tion I have given above ; in these Arthropods at least yelk 

 segmentation is holoblastic, and the yelk grows after the morula 

 is formed. So also in Peripatus, if the observations of Kennel 

 are to be trusted and they certainly appear to be far more 

 trustworthy than those made by his opponents the yelk seg- 

 mentation in this group is holoblastic. 



In conclusion, I would observe that although the older 

 investigators held that the blastoderm in Insects appears 

 simultaneously over the whole surface of the yelk, more recent 



