532 THE FUR SEALS OF THE PRIBILOF ISLANDS. 



Both Synoieum irregulare and PolycUnum pannosum present interesting conditions 

 in connection with their sexnal reproduction. I describe that shown by the first- 

 named species only. Unfortunately, however, the collection does not contain sufficient 

 specimens to enable me to make the account as full as might be desired. 



In the formal descrijition of the species I have pointed out that the thorax is 

 very small relatively, and is much contracted and so dense and opaque as to make it 

 impossible to distinguish the branchial sac with any clearness. This is the condition 

 in all the colonies at hand in which the thorax of the zooids is found at all. But in 

 most of the colonies a great proportion of the zooids are wholly without the thorax. 

 It frequently happens that, although the lobes of the colonies are of nearly noi-mal 

 size, the upper half or more of some of them may contain no zooids or parts of zooids, 

 and the test may be entirely solid, i. e., without cavities such as are usually found in 

 preserved specimens of compound ascidians in which the zooids have withdrawn 

 upon killing into the deeper portions of the test. This condition is the result of 

 degeneration of the zooids, or at least their anterior portions. Post-abdomens alone 

 are found in great numbers in such colonies. Freciuently these do not differ in any 

 respect, either in form, size, or composition, from others that are still connected with 

 branchial sacs. I have not been able to follow the process of disappearance, either 

 of the thorax or of the solidification of the test in the i)arts of the lobes containing 

 no zooids. It is very ])ossible that the condition of the thorax as I have described it 

 is not typical for the species, but is a result of the retrogressive process having already 

 set in. Of this, however, I have no certain proof. Examination of the free post- 

 abdomens shows them to be densely filled with a material that is for the most part 

 undoubtedly of the nature of food yolk. Tliis does not differ in any essential parti(!- 

 ular from the food material that is found in many compound ascidians. When Cully 

 elaborated it consists of an enormous number of small granules, very regular in size, 

 form, and oi)ti('al i)roperties. They are almost perfectly spherical and are highly 

 refractive, and possess a slightly yellowish tinge. 



In many species these granules can be easily seen to be contained in the mesen- 

 chymatous cells, but here no evidence of cell structuie in connection with them can 

 be made out in most cases. It is i)robable that the cell substance has become wholly 

 converted into the granules, though just how the thing is accomplished is not clear, 

 since the bodies into which the granules are aggregated are much larger than the 

 individual mesenchyme or body-substance cells ever are, and at the same time they 

 appear to be too definite and constant in form to permit, without much misgiving, the 

 sui)position that they are formed from the running together of several cells. Their 

 form approaches spherical in almost ail cases where they are not under external jires- 

 sure from some cause. Many of them reach a diameter of 45 or 50 //, while their aver- 

 age size would ])r()bably be about 30 /y. I<>oni their form and behavior under pressure 

 of the cover slip, and from what is known of the similar bodies in other species, it is 

 quite certain that each one possesses an exceedingly thin membrane. But this is 

 difficult to prove directly. In most, if not in all of these free post-abdomens jmrtions 

 of the mantle containing the characteristic muscle fibers and epithelial cells are 

 present, and also the heart and the epicardiac tubes may frequently be found. 



But the most interesting facts in connection with them relate to the sexual cells 

 and their development. In many of the abdomens, particularly those that are least 

 changed in form and structure, the band-shaped ovary is found to differ in no respect 



