280 TSIR C. ELIOT ON NTIDIBRAXCHS [NoV. 29, 



Phyllidiopsis and Ceratophylliclia the mouth-pai'ts are much as 

 in Doridopsis, the glands not being fused with the buccal tube, 

 iind CeratophyUidia has the additional peculiarity of bearing 

 stalked globes on the back. The distinction between Phyllidia 

 and Phyllidiella seems to me less certain. According to Bergh, 

 («) the oiul tube is symmetrical in Phyllidia^ asymmetrical in 

 Phyllidiella ; but I have not found the diflerence to be clear or 

 persistent, and even if it is so, I doubt if it is of generic woith. 

 \h) In Phyllidia, " Dorsum tuberculis elongatis, plus minusve con- 

 fluentibus obsitimi, medio varicositates longitudinales formant- 

 ibus." In Phyllidiella, " Dorsum proprium tuberculis discretis 

 vel pro parte confluentibus quincunces formantibus obsitum." 

 Even in typical forms it does not appear that this distinction is 

 clear. Phyllidia elegans (see Bergh's jVIonogi^aph, pi. xix. fig. 1) 

 seems to me to have not " varicositates longitudinales," but 

 gi'oups of confluent tubercles ; and, on the other hand, Phyllidiella 

 pusiulosa strikes me as having not so much tubercles arranged in 

 *' quincunces," as compound tubercles arranged in lines. But 

 in abnormal foi"ms, which are frequent, it is still harder to draw 

 the distinction. I have a fine specimen with all the characters of 

 Phyllidia varicosa, but the median iddges, though very distinct 

 posteriorly, are bi'oken up in front and unite to form a quincunx 

 iis in PhyUidiella nobilis. Again, in this latter, the quincunces 

 Are often placed so I'egularly above one another that the tubercles 

 ■seem to be ari-anged in longitudinal lines anei not in figures. 

 I therefoi'e think it better to abandon PhyUidiella as a separate 

 jgenus, 



Tlie ai'rangement of the dorsal tiibei-cles in these foi-ms is so 

 variable, that it is hard to draw the line between species and 

 varieties, but at least three certainly specific forms occur in East 

 Africa : — 



(1) Ph. varicosa. — Coloui- black, blue, and orange ; rhinophores 

 yellow. A black stripe on the foot. Tubercles more or less 

 fused into ridges, but not compound. Tj^pically, thei-e are three 

 long ridges down the centre of the back and a number of short 

 ridges, more oi' less at right angles to them, running inwards from 

 the mantle-edge. Buccal mass yellow and very complicated. 



(2) Ph. pustulosa. — Black and gi'een ; rhinophores black, no 

 black line on foot. Tubercles simple or composed of only two or 

 three lumps. Buccal bulb partly black. 



» (3) Ph. nohilis. — Coloui- as in Ph. pustidosa, but tubercles 

 highly compound, sometimes consisting of ten small lumps fused 

 together. Typically, they ai'e arranged in square or oblong figure 

 Buccal bulb lai-ge, yellow. 



That the pattei-ns foi-med by the tubercles should vary is not 

 surprising, if we consider that the tubei'cles have always a 

 tendency to unite, and may do so more or less decidedly in a, 

 given dii'ection. Thus the typical form of Ph. i-aricosa occurs 

 when the lateral tubercles unite in a predominantly transverse 

 ^lirection. When however, union in a longitudinal direction 



