268 McMURRICH. [Vol. II. 



IV. The Development of the Filaments in Buds. 



The time relations of the cihated bands and glandular streaks 

 in buds is just the reverse of what obtains in egg embryos ; that 

 is to say, the ciliated bands are the first to develop, the glandu- 

 lar streaks appearing later. 



In a bud of Z. sociatiis 2 mm. in length the stomatodaeum 

 is already formed, and on the edges of the perfect mesenteries, 

 immediately below the lower margin of the stomatodaeum, the 

 ciliated bands can be seen presenting practically the same 

 appearance as in adult polyps. Following a band downwards, 

 it is found to disappear below and no trace of a glandular streak 

 can be found, and no enlargement of the endoderm just exter- 

 nal to the free edge of the mesentery. Indeed, there is noth- 

 ing to distinguish a perfect mesentery from an imperfect one 

 below the level of the stomatodaeum, except greater width. 

 The free edges of both mesenteries is occupied by cells 

 indistinguishable from ordinary endoderm except by their 

 apparently somewhat smaller size. 



The glandular streak begins to develop, however, soon after 

 this stage, since in a bud but little older they were readily 

 recognizable, and the ectoderm just external to them had 

 become relatively high and was packed with foreign bodies ; in 

 buds 3.5 mm. in length all the parts of the filament occurring 

 in the adult were present. 



It is interesting to note that in the buds of Alcyonaria the 

 same acceleration in the development of the ciliated bands has 

 been observed by E. B. Wilson ('84), the glandular streaks in 

 the egg embryos of these forms developing before the ciliated 

 bands, as in zoanthids. 



V. Conclusion. 



I have shown above (i) that in adult polyps of Z. sociatiis 

 there is no histological continuity between the glandular streaks 

 and the ciliated bands ; (2) that in Q.gg embryos the glandular 

 streaks develop before the ciliated bands make their appear- 



