ELIOT: NUDIBRANCHS FRO^r THE INPO-PACIFIC. 245 



say, "No specimen of this very distinct species was in the collection," 

 from which it may be inferred that their description was made entirely 

 from the drawing, so that mistakes may have occurred. The original 

 drawing must have been better than my copy of the printed plate if 

 it enabled them to see all the details they record. 



DORIDID.« CRYPTOBRANCHIAT^. 



Chromodoris Alder & Hancock. 



About 130 species of this genus have been described, mostly from 

 external characters only. Further examination and particularly a 

 comparison of living specimens will probably show that this number 

 ought to be considerably reduced. The best specific distinctions 

 seem to be those afforded by the radula and labial armature, although 

 these certainly vary in size according to the age of the individual and 

 perhaps in other respects. The coloration and pattern are strikingly 

 varying, and animals of widely different appearance are found to be 

 connected by intermediate forms. It would seem that the human 

 eye is not quite impartial in estimating variations of colour. As long 

 as the coloration is subdued and the pattern indefinite, considerable 

 variation is accepted without question within the limits of a species. 

 A striking example of this is Archidoris iubenulata^ which ranges from 

 clear pale yellow to dark brown, with or without mottlings of grey, 

 green, or pink. But when the coloration is vivid and the pattern 

 distinct and symmetrical there is a strong tendency to create new 

 species, although the amount of variation may not be greater than 

 that exhibited in A. tuberailafa. 



Of the species described here, Chr. saniiio seems the most distinct, 

 having tubercles on the back (an unusual character in this genus) and 

 spherical glands below the mantle edge. Chr. i'^////^;'/ appears to be 

 connected by clear intermediate links with Chr. nigrostriata and Chr. 

 tenuiliiiearis, which must hence be regarded as varieties. This is a 

 very interesting case of variation, because it shows how remarkably 

 distinct lines, which appear specifically characteristic, can be broken 

 up into spots and dots, producing a pattern in which there are no 

 lines at all. 



Chr. pttsilla (hitherto known only by a figure) belongs to a group 

 characterized externally by having a few longitudinal lines on the back, 

 and internally by having bifid teeth, with a few denticles below the 

 bifurcation. Chr. liiieata, Chr. hilaris, and probably Chr. niagnifica 

 all belong to this group, and are perhaps merely varieties of one 

 species. 



Chr. petechialis belongs to a group characterized externally by a 

 coloured border round the mantle, and spots on the dor.sal surface. 

 The teeth are not bifid but denticulate and rather erect. Chr. 



