102 SIR U. ELIOT ON XUDIBRANCnS [May 17, 



rhinopliore ; the fii'st pair of cerata receive no branches at all ; 

 the second and third receive on the right hand a branch which 

 bifurcates, and on the left a simple branch which, in the third, 

 stops at the base of the cera and does not enter it. The re- 

 maining cerata receive no branches. 



I think these specimens are the B. digitata described by A. & H. 

 and by Bergh. The best external character seems to be the 

 tapering, finger-like shape of the cerata and of the process behind 

 the rhinophores, to which no doubt the specific name is due. 



BORNELLA ARBORESCEXS Pease. 



[Bergh, "Neue ISTacktschnecken, 'No. ii.," Journ. Mus. Godeffroy, 

 Heft vi. 1874, p. 96; id. S. R. xvii. p. 886.] 



Several sjiecimens from Mombasa Harbour. Note on living 

 animals : " Yellowish white, with red reticulations on back and red 

 tips to cei-ata." 



The alcoholic specimens are all much of the same size. All are 

 whiter and more compressed than tliose of B. digitata, and the 

 cerata are much smaller. The average dimensions are : — Length 

 20 millimetres, height 6, breadth 3 ; rhinophores and cerata 

 about 2 mm. high. The tentacular processes at the side of the 

 mouth consist of only about six digitations. The rhinophores are 

 as in B. digitata, but the posterior process is not so long. In most 

 specimens there are five paii-s of cerata, of which the fii-st three 

 ai'e bifid and the remaining two simple. Each, from the first to 

 the fourth, bears two gills, the fifth none. The jaws and labial 

 armature are as in B. digitata ; the formula of the radula is 

 about 40x9.1.9, rising sometimes to 12.1.12. The teeth are 

 niuch as in Bergh's plates (Journ. IVIus. Godef . I. c. plate iv. 1 2), 

 but the central cusp of the median tooth is I'ather longer. The 

 median tooth is more erect than in B. digitata, and the 8-10 

 denticles which it bears less distinct and very hard to see. The 

 other characters are as in B. digitata. 



It is not easy to say whether this form is specifically distinct 

 from B. digitata or, if so, whether it should be called B. arbo- 

 rescens. It represents, however, at least a well-marked variety or 

 stage of growth in which the tentacular processes, rhinophores, 

 and cerata are less amply developed. It could hardly he identified 

 with B. arhorescens on the strength of the original description by 

 Pease (Amer. Journ. of Conchol. vi. 1871), but in the revised 

 description by Bergh (Mus. Godef. I. c.) the chief specific character 

 seems to be '^papillis anteriorihus lot plurimum bipartitis." In 

 these specimens they are invariably bifid. With regard to the 

 hooks on the prfeputium, I was unable to see the difference men- 

 tioned by Bergh, and found only simple or bifid hooks, not trifid, 

 in both species. 



BoRXELLA EXCEPTA B. (Plate lY. fig. 5 b.) 



[Bei'gh, Challenger Reports, Nudibranchiata, p. 36 ; id. Danish 

 Exp, to 8iam, Opisthobranchiata, p. 202.] 



[20] 



