534 THE FUR SEALS OF THE PEIBILOP ISLANDS. 



comparatively late stage in tlie growth of the latter. If such be the case, it is probably 

 due to the fact that the young ova and the "test" and follicular epithelial cells 

 furnish a more accessible and an ample food supply for the ova during the early 

 stages of their growth. This would result not only from the fact that the ova are 

 held in the ovary for a time, but also from the farther fact that the masses of yolk 

 granules are, as already pointed out, enveloped by a membrane at the time when the 

 post-abdomen is set free from the zooid. 



It is worthy of special notice in this connection that at no time in the career of 

 the growing ovarian ova are there, so far as my observations have gone, any indica- 

 tions of amoeboid or other changes in their nuclei. As is seen by reference to figure 

 30, PI. LXXXVI, the germinative vesicle presents in each ovum the familiar charac- 

 teristics of this body in ovarian ova, and this notwithstanding the fact that the 

 ova are actively ingesting and presumably digesting also. 



The embryonic stages which I have observed are the early cleavage stages 

 (PI. LXXXVI, flg. 31) ; late morula and early gastrula stages (fig. 32) ; fully developed 

 tadpoles, and tadpoles in which the metamorphosis is well advanced (fig. 33). I 

 describe the last two of these first. On making a section of the lobes of several of the 

 colonies, packets of bodies, a few of which are shown in figure 27, are found embedded 

 in the semi cartilaginous test. Cursory examination proves the bodies to be embryos 

 iu various stages of development. In one capsule, almost perfectly spherical, 3 mm. in 

 diameter, were contained 13 embryos; in another, 3.5 mm. in diameter, were 16 

 embryos. Others examined contained fewer than the first mentioned, but none more 

 than the last. The embryos are very closely packed together in the capsule, and they 

 constitute its entire contents, so that after they are picked out the capsule is entirely 

 empty and its interior is almost as regular and smooth as that of a bullet mold. 

 The capsules are perfectly closed at all points. Concerning the embryos themselves, 

 not much need be said. Figure 33 shows one in which the metamorphosis is well 

 advanced. 



The structural fact of most interest in connection with the full-grown embryos is 

 the thickness and the composition of their own test. A general idea of this is giveu 

 by figure 33 ts (PI. LXXXVI), and a more detailed representation is shown by figure 

 34. The interest that attaches to this point lies in the probable fact that some of the 

 elements contained in the test are unconsumed remnants of the extra ovarian portions 

 of the post- abdomens of the parent zooids. There can be scarcely a doubt that such 

 is the nature of the bodies shown at y% figure 34. These have the form, size, and 

 composition of many of the masses of yolk granules already described as constituting 

 so large a part of the bulk of the recently severed post-abdomen. In another part of 

 the test of the same embryo there occurred a considerable number of fibers (PI. 

 LXXXVI, fig. 34 m. /.) which so strongly resembled the ordinai-y muscle fibers of the 

 mantle of the adult that I should not have thought of questioning their nature but for 

 the remarkable position occuiiied by them. Structures more or less similar both to the 

 masses and the fibers are very common in almost every embryo. Most of these certainly 

 belong to the test of the embryo itself, and are of course the same as those Similarly 

 situated in the embryos of all tunicates. But the number is here unusually large, and 

 when this circumstance is considered along with that of the structure of the ones 

 described above, it appears almost certain that, as already said, some of the various 

 bodies contained in the test of the embryos are remnants of the parent zooids. Whetlier 



