1918.] J. Stephenson: Oligocliaetn of the Inle Lake. 15 



of the conjoined atrium and paratrium ; the evagination of the terminal 

 section of the atrium seems never to be complete, so that there is in 

 full protrusion still a deep groove round the base of the penis, which 

 thus projects from within the male aperture (text-fig. 3). 



I was at one time inclined to doubt whether the " coelomic sac '" 

 was really coelomic, — whether its cavity was really a cut-off portion of 

 the coelom. It seemed to me that the most terminal portion of the male 

 duct, — that part which was originally included in the parietes, — might 

 have hypertrophied to produce the penis, and in so doing might have 

 raised up the inner part of the muscular layer of the bodywall so as to 

 form the sac. On this supposition the cavity of the sac would not be 

 coelomic, but only an enlarged split in the muscular wall of the body. 

 In the same way the cavity of the gills is a space, not coelomic in origin, 

 between the two muscular layers of the parietes ; there the outer muscular 

 layer carrying the superficial epithelium projects outwards as a gill, here 

 the inner layer carrying the peritoneal investment would project inwards 

 as the sac-wall. 



However, the atrium within the sac has an (apparently) peritoneal 

 covering, well-marked in places, and consisting of cubical clear cells ; 

 and a much flattened cell-layer can also be seen on the inner side of the 

 sac- wall. The cavity and its contents appear therefore to be lined and 

 covered respectively by peritoneal epithelium, and the space to be really 

 coelomic. 



The above condition is remarkably similar to that which I have 

 recently described in Kawamuria (11). It is curious that the penial pro- 

 jection in B. sowerhyi has not previously been observed ; there is how- 

 ever nothing, in the more usual condition of the orifice, to indicate the 

 possibility of such a protrusion, — the canal ends quite simply on the 

 surface of the body. It is true that I recognized the function of the 

 coelomic sac in Kawamnria, — to cause, by contraction of its walls, the 

 extroversion of the contained tube ; brt in Branchiura, which has the 

 sac but, so far as I then knew, no protrusible penis, I considered the sac 

 to have lost its function, and to be a rudimentary organ ; it is evidently 

 at times fully functional. 



The distance between Branchiura and Kawamuria is thus reduced : 

 the separation must now depend on the presence or absence of gills. I 

 will not further discuss at present whether generic distinction is still 

 justifiable, but may refer to what I wrote in my former paper. 



The specimens showed much variation. In length one batch con- 

 sisted of worms of 30, 28, 25 or fewer mm. ; in another the individuals 

 were 45—70 mm. ; in others they were 90 or 100 mm., and one specimen 

 reached the great length of 185 mm. 



The number of pairs of gills was 40—47 in the shorter worms, and 90 

 in the longest ; but this was not the maximum. In one fragment 140 

 pairs were found, but the total length of the animal cannot be known. 

 As many as 110 were found in an individual only 70 mm. long. 



Sometimes nearly all the gills were well developed, only a few at 

 the anterior end of the series being represented by mere tubercles ; in 

 others a large number of the anterior gills were only tiny projections. 

 But variation in this point seems to have no relation to the number of 



C 



