334 E. W. MACBRIDB. 



The second type of embryonic development, viz. that in 

 which the egg is applied to the uterine wall, is characterised 

 by the reduction of the food yolk, so that the segmentation 

 reverts to the total type. Sharply marked traces, however, of 

 the former presence of yolk remain. The gorging of the en- 

 doderm with yolk has rendered the archenteron functionless, 

 and as it still remains functionless when the plan of absorbing 

 nutriment from the uterus through the general surface is 

 introduced it is deferred in development, and in fact the 

 destiny of the first products of segmentation is totally different 

 in Mammalia from what it is, for example, in Echinoderms. 



In the third type of embryonic development, in which the 

 ovum is enclosed in a capsule with a number of yolk cells, the 

 most weird changes are produced in development. We read 

 of a complete separation and subsequent reunion of the blasto- 

 meres for instance ; this type is almost confined to the 

 Platyhelminths, and is alluded to here only for the sake of 

 completeness, and to show how few traces of ancestral history 

 the development of these animals affords. One family only, 

 the Dendrocoela, lay their eggs singly, and in this case we have 

 a large amount of food yolk present ; yet Platyhelminths 

 have bulked largely in many phylogenetic speculations. 



Lastly, we have those few cases in which the developing 

 animal escapes from the egg-membrane, but remains in the 

 uterus or brood-pouch. In these cases we have comparatively 

 little interference with the normal course of development. The 

 early stages occur in a perfectly regular manner, and we have, 

 in fact, free-swimming larvae within the brood-pouch ; it is 

 only in the later stages that they commence to absorb fluid 

 from its walls. It is necessary to emphasise this type of 

 development, though it is comparatively rare (Brachiopods, 

 Paludina, and Amphiura squamata) because if it is con- 

 founded with the foregoing types, its totally different character- 

 istics would seem quite inexplicable. 



I ought perhaps to mention that the earlier stages of 

 Amphiura squamata are described as being abnormal by 



