392 W. G. Van Name — Bermuda Ascidians. 



ries in vaiying numbers, and communicate Avitb the sperm duct 

 which follows the ovary by slender connecting- ducts. Usually the 

 larger the individual the more numerous and more extensively 

 branched are the testes, though this is not always the case. 



After this description of the New England form it Avill be suffi- 

 cient to mention the particulars in Avbich the Bermuda variety differs 

 from it. 



In the first place, it is of considerably smaller size, the largest 

 specimen obtained measuring 22™"" by 10™™. Most of them were 

 hardly more than half this size. It is not unlikely, however, that if 

 the collections had been made later in the season, larger specimens 

 might have been found. 



In external form it appears to vary more than the t^^pical jyartita. 

 It is attached either by a small area near the posterior end or by a 

 part or the whole of the ventral surface, and in the latter case the 

 siphons are both situated on the dorsal surface. The character of 

 the surface of the test is very variable ; it is generally roughest 

 near the apertures, which are usually more or less prominent, but 

 whether the ridges and wrinkles of the surface are large or small, 

 regularly or irregularly disposed, low and rounded or prominent and 

 sharp-edged, appears to be a character of no specific value. 



The color is generally a more or less reddish or brownish yellow, or 

 grayish j^ellow, becoming brown or red on the upper surface, espe- 

 cially about the siphons. The colors are brighter and the test pro- 

 portionately thicker and of a more cartilaginous character than in 

 the New England specimens. The striping of the apertures, which 

 many specimens show in common with the typical 2yaTt'ita, has been 

 mentioned above. 



As figs. 71 to 75 indicate, the form of the body and length and 

 position of the siphons are very variable. The mantle is thinner, 

 less muscular, and more transjiarent, though of a deeper yellow color 

 in most cases, and the tentacles are rather more numerous, but the 

 branchial sac does not appear to differ essentially from that of New 

 England specimens of similar size. 



The more usual form of the orifice of the dorsal tubercle is a U 

 or horseshoe-shape, with one horn curved inward and posteriorly, 

 alongside the other, but not spirally coiled. Considerably more 

 comjjlex forms occur, as is also the case in the true partita. 



The reproductive organs are similar, but the testes are fewer and 

 often are not branched at all, but merely simple elongated bodies. 



This form was found in many localities about the islands and on 



