MUSEUM OF COMPARATIVE ZOOLOGY. 265 



have a close resemblance to certain similarly placed threads in worm 

 larvae, as in the well-known Tornaria, where they extend from the cavity 

 of the larva to an ajiical cluster of modified epiblastic cells through an 

 intermediate gelatinous layer. I was unable to observe these threads 

 closely enough to detect any tubular structure in them. Later in the 

 growth of the larva there are two filiform bodies connecting the cavity 

 of a hydrophyllium with its surface, which may possibly be the same as 

 the thread-like extensions of which we have already spoken. In the 

 development of Agalma Sarsii as figured by IMetschnikoff, we have no 

 representative of these threads in the primitive larva, or in stages of 

 later growth. Ha^ckel figures certain structures in the hydrophyllium 

 of Physophora which have been found by me in Agalma, which in P/ii/- 

 sophora have the form of small tubes extending from the cavity to the 

 surface. These call to mind the protoplasmic bodies in the piimitive 

 hydrophyllium of Agalma, although they are different in many respects. 

 Hseckel gives them a morphological significance in Physophora, and 

 regards them as comparable with certain parts of the chymiferous ra- 

 dial tube system of hydroid gonophores. He does not represent them 

 in the younger forms, at least, of the primitive larva of Crystallodes. 

 Cilia were not observed on the outer surface of the primitive covering- 

 scale, but were seen on the epiblast covering the yolk at this agei 



Of the remaining structures found in the primitive larva the most im- 

 portant in the future history is a spherical organ {pii. cy.) adjacent to the 

 end of the primitive cavity. This body is the future float, and at this 

 stage lies inside the egg, or between the yolk cells and the superficial 

 covering, although no mai'ked external elevation could be seen. The 

 float is enclosed by a layer of cells which was traced continuously into 

 the hypoblast of the primitive cavity, and also into the hypoblast 

 which covers the yolk sac. Within the hypoblast the contents of the 

 float and the hypoblastic layer were slightly separated. A continuation 

 of the same layer, epiblast, reflexed from the inner surface of the cover- 

 ing-scale, extends over the float and is continued over the surface of 

 the egg. 



A second appendage, which assumed the form of a slight projection 

 from the surface of the yolk on the left-hand side of the cavity of the 

 primitive hydrophyllium, is also present in this stage of the primitive 

 larva. In profile this structure (ser. hyph.) is arch-shaped, and has 

 a slightly reddish color. It is the beginning of a covering-scale which, 

 although provisional in nature, has given the name of "Athorybia 

 stage " to a larval condition of Agalma which follows the first or primi- 



