488 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. 45. 



Second Loring collection, 5 specimens, size 2 by 1^ cm., test very 

 thin, soft, and transparent; lobes of orifices uncertain, 6, 5, or 4, 

 tentacles 43 to 45; dorsal languets large and well separated; internal 

 longitudinal vessels of sac about 22, and small papillae on them cor- 

 responding to the short vessels by which the longitudinal vessels are 

 connected with the branchial membrane; stigmatic spirals very 

 regular, as in Puget Sound specimens; stomach globular, wall with 

 20 to 25 folds. 



All the Loring specimens were attached to Halocynihia villosa, and 

 were elongate and narrowed toward the point of attachment. 



Puget Sound specimens from O. B. Johnson, July, 1889. 



Loring specimens. Albatross, Alaska salmon investigations, April 29, 

 1903. 



Note. — Huntsman's 0. rugosa, the description of which has come 

 to hand since the above was written, does not do away with the 

 variational difficulties noted by me. The roughened test and more 

 anterior position of the atrial aperture which, according to Huntsman, 

 distinguish rugosa from willmeriana are certainly inapplicable as dis- 

 tinctive marks for the specimens before me, both being so far as I 

 can see strictly individually and quite independently variable. My 

 largest Puget Sound individual is entirely smooth-surfaced, while a 

 second, nearly as large, is unmistakably roughened and papillate. 

 A smaller individual is smooth or nearly so on one side and conspicu- 

 ously rough on the other. The Loring specimens present much the 

 same range of variation in this particular. 



The other point on which Huntsman relies for separating the species 

 is the number of longitudinal vessels, willmeriana being assigned 24 

 right and 22 left; while rugosa is given 20 to 22 on each side. This 

 difference is quite too small to be held as specifically significant unless 

 based on averages of a large number of determinations. As a matter 

 of fact I find 20 on each side in a large Puget Sound individual, and 

 22 to 23 in the Loring animals. 



CORELLA JAPONICA Herdman. 



Plate 35, figs. 28-30. 



Corella japonica Herdman, 1882, p. 190, pi. 16, figs. 1-9.— Sluiter, 1900, p. 20 — 

 Hartmeyer, 1906, p. 25; 1909a, p. 1393. 



Discussing this species in his final report on the Challenger ascidians 

 Herdman recognized its close resemblance to C. eumyota Traustedt 

 (1882), but pointed out a number of characteristics which seemed 

 sufficient to distinguish it from Traustedt's species. The results 

 of my comparison of the specimens at hand with the descriptions of 

 jajjonica and eumyota led me at first to the conclusion tliat my speci- 

 mens, even though from the Japanese coast, agreed more closely 

 with eumyota than with japonica. This was surprising because 



