524 THE FUR SEALS OF THE PRIBILOF ISLANDS. 



more ox* less prououuced and regular tbickeiied areas in the wall, and m connection 

 with these, at least in many of the preserved s[)ecimeus, there are indications of 

 folding, (Fig. 20.) But in some specimens, again, I can detect scarcely a trace of 

 either thickening or folding, and there are so many and close resemblances between 

 our species and ^i. mrsii, Huitfeldt-Kaas, 1890, from the Norwegian coast, that I am 

 fully convinced of the very close affinities of the two. I have consecpiently deemed 

 it the wiser course to place it here rather than to add another to the already long list 

 of rather illy-defined genera into which the Polyclinidae are divided, lu this connec- 

 tion I can not refrain, after having spent much time in examining the stomachs of 

 numerous species and genera, and in critically reading the utterances of other writers, 

 from quoting that master zoologist, Milne-Edwards, 1842, on this point. After 

 describing the stomach of his Amaroucium argus (transferred by Giard, 1872, largely 

 on account of the structure of the stomach, to the genus MorchclUum),he says: "Mais 

 si V on dcscendait a des caracteres de cet ordre pour en faire la base des divisions 

 gcncriques, on serait conduit ;\ multiplier inutileinent ces coupes et on reudrait les 

 determinations d' une difficulte extreme." The advance of knowledge since this 

 remark was made has undoubtedly shown that the character of the stomach wall is 

 of diagnostic importance, but that it in itself can be relied upon in all cases, even as 

 a distinctive S2)ecific mark, to say nothing of its generic value, I do not believe. The 

 si)ecies is re])resented in the collection by a single colony from St. Paul Island. 



Amaroucium kiucaidi, new spocics. 



Colony cake-like, irregular in outline, but always depressed. When attached to 

 cylindrical bodies of small diameter, as seaweeds, which seems to be the usual habit, 

 colony entirely incircling these. Firm and unyielding to the touch. In color, little 

 sand on the surface, zooids showing through the test (piite distinctly. Common atrial 

 orifices large and open. 



Dimensions of largest colony, 54 mm. to 31 mm. to 21 nun.; of smallest colony, 

 27 mm. to 20 mm. to 17 mm. 



Test — Surface layer, which is (^uite thick, containing so much brownish-gray 

 coloring matter that the transparency characteristic of the interior portions is here 

 wholly obscured. Cells of the test vary; abundant fibers absent. Sand grains 

 penetrating the entire mass, though not numerous, not incrusting the surface. Only 

 a small interior portion of test into which tlie zooids do not reach. 



Mantle. — Longitudinal muscle fibers in bands, though not as comi)letely separated 

 fronx one another as in many species. Circular fibers almost wholly absent excepting 

 in the siphons, but not numerous even here. 



Zooids. — Visible, though not distinctly so, on the surface of the colony. Arranged 

 in definite, circular systems, each containing about six or eight individuals. Standing 

 at various angles to tire surface of the colony, and considerably contorted, so that 

 they appear only in fragments on cut surfaces of the colony, whatever be the direction 

 of the section. Subdivisions of the body not distinct. Size, medium; total length, 

 about 8 mm.; thorax, about 2.5 mm.; abdomen, about 1.5 mm., and post-abdomen 

 about 4 mm. Peduncular i>ortion of post-abdomen easily noticeable, though not 

 conspicuous as compared with many other species (fig. 21). An ectodermal appendage 

 of the mantle at the posterior end of the post-abdomen. 



Branchial apparatus. — Branchial orifice scarcely recognizable on the surface of 

 the colony. Common atrial orifice, large and open, at least in some colonies. 



