TUmCATA OF THE PRIBILOF ISLANDS. 533 



from the condition which it presents in ordinary normal zooids. (Fig. 28, ov.) The ova, 

 however, many of them at least, are distinctly amoeboid, and at no period of their exist- 

 ence are they enveloped by either a follicular epithelium or "test" cells (PI. LXXXVI, 

 flgs. 29, 29", 29*.) The absence of these two layers, the latter of which in particular 

 is so characteristic of the tunicate ovum, is noteworthy. I know of no other instance 

 of the kind in the group. This peculiarity of the individual ova gives an appearance 

 to the ovary as a whole strikingly different from that of the ordinary ovary of these 

 animals. The ova are very closely packed together, and the pseudopodia-like proc- 

 esses, in some instances quite long and narrow, lock together and overlap in an 

 intricate way; and as the cytoplasm is quite homogeneous and refractile, particularly 

 in the smaller and middle-sized ova, the appearance is, as remarked above, striking. 

 Whether or not this amoeboid condition prevails before the abdomen is separated from 

 the rest of the zooid, I do not know; but in all probability it does, since the smaller 

 ova of the severed abdomens show it to almost as great a degree as do the larger 

 ones. The character is, however, wholly lost before maturation takes place. At 

 least this is the case so far as my observations have gone. I have found a few ova, 

 one of which is shown in PI. LXXXVI, fig. 30, that are perfectly spherical, and as the 

 cytoplasm of these is entirely filled with food granules, I assume that they are nearly 

 ready to undergo the maturation changes. 



In addition to the amoeboid form of the ova, they show the same nature to a still 

 greater extent in their j)ower of ingesting other cells. Figs. 29 and 29* (PI. LXXXVI) 

 illustrate this. That the small cells are actually contained in the cytoplasm of the ova, 

 and are not merely situated on the surface, may be shown conclusively by isolating the 

 ova and so manipulating the cover slip as to cause them to swim about and turn over in 

 the fluid in which they are contained. Such ova as the ones figured, showing the cells 

 in various stages of penetration and disintegration, are very abundant. Ova are easily 

 found in which as many as five or six of the ingested cells may be seen. 



I have not been able to satisfy myself as to the nature of these cells. Such 

 instances as that shown at a, fig. 29*, where the cell is only embedded in the surface 

 of the ovum, gives rise to the suspicion that they represent either the follicular 

 epithelial cells or the " test " cells, characteristic of the ova of tunicates. They may 

 also, at least in some cases, be very young ova. Indeed it is highly probable that 

 many of the ova are consumed iy their companions, for certain it is that only a small 

 fraction of the entire number contained in an ovary ever develop into embryos. I have 

 said that the cytoplasm of the smaller and middle-sized ova is quite homogeneous 

 and refractile, and also that in the older ones it is filled with food granules. These 

 granules in such an ovum as the one shown in fig. 30, for example, are not recognizably 

 different from those already described as constituting most of the bulk of the large 

 bodies which I have said fill the post-abdomen, and which are in all probability yolk- 

 laden mesenchyme cells. 



There is little doubt that the yolk granules serve as nutriment for the growimg 

 ova and embryos, as do the ingested young ova and " test" cells ( ?). It is true I have 

 not been able to actually observe the ingestion of the granules by the ova, but the 

 fact that they have wholly disappeared from the cavities in which the embryos are 

 situated by the time the fully developed tadpole stage is reached hardly adnjits of 

 any other explanation. My failure to observe the ingestion of the granules by the 

 growing ova may be dlie to the fact that the process actually does not begin until a 



