430 PEECT SLADEN TRUST EXPEDITION. 



dark brown in front, yellow behind. There are about 80 rows, of which the longest con- 

 tain 90-100 teeth on either side of the rhachis, which bears slight triangular thickenings. 

 The teeth are strong, erect, and hamate, not bilid but bearing as a rule 6-9 denticles 

 rather low down on the shaft. The denticles are short, thick, and not conspicuous, 

 though when the teeth are in a favourable position they can be seen quite clearly. As 

 in most Chromodorids, the innermost teeth are broad and denticulate on both sides ; the 

 outer teeth are straighter, less hamate than the others, and bear denticles on or near 



the tip. 



The other internal organs appear to be as usual in the genus. The oesophagus runs 

 straight into the liver without any dilatation, and the hepatic cavity acts as a stomach. 

 Undigested food was found in two chambers within the hepatic mass. 



There appear to be two variable species of Chromodoris which in certain varieties 

 present a similar external appearance. One of these may be called Chromodoris quadri- 

 color (Riippell), which was admitted by Bergh to be the same as his Chr. elizabethina and 

 Chr. anncB. In the typical form this species is characterized externally by stripes of four 

 colours — white, orange (red or yellow), dark blue, and light blue. But the dark and light 

 blue may both become black and the relative quantity of dark and light colour may vary, 

 so that some specimens appear to be dark animals with light stripes, others to bear dark 

 stripes on a light ground. In all these varieties the radula seems to be similar. 

 It consists of rather erect teeth, not bifid at the tips, and bearing about eight 

 denticles. 



The other species may be provisionally called Chr. lineata (Souleyet, 1852) and is pro- 

 bably identical with Chr. magnifica, Q. & G. (1832), and Chr. hilaris, Bergh. These 

 forms are yellowish, with orange-coloured branchiae and rhinophores and violet striijes on 

 the back and sides. The teeth of the radula are bifid (so far as is known) and bear about 

 four small denticles below the bifurcation. 



It will be seen that if the stripes of Chr. quadricolor are multiplied, it becomes 

 similar both in pattern and coloration to Chr. lineata or Chr. magnifica. This has 

 happened in the specimen here examined, but as it has teeth of the shape found 

 in the typical form of Chr. quadricolor, I think it should be referred to that 

 species. 



It may be asked if the teeth are not as variable as the coloration. It may be so, but 

 the specimens of Chr. quadricolor hitherto examined, though differing in external 

 appearance, present the same type of tooth. It is possible that Chr. quadricolor and 

 Chr. linearis may both prove to be varieties of one widespread and very variable species, 

 but at present I think nothing would be gained by ignoring the differences which seem 

 to separate them. Though the specific name magnifica has priority, it is better not to 

 use it until the character of the radula has been ascertained in specimens indubitably 

 referable to the species. Both the present specimen and the one from Zanzibar 

 described by me {I. c.) resemble Quoy and Gaimard's plate, but their radulse are not 

 the same 



