224 BULLETIlSr 15 4, UNITED STATES NATIONAL MUSEUM 



The six i'onus oi" the te^scllatas <^vo\\\) thai, have been found to 

 occur exclusively on ishinds in the Gulf of California are clearly 

 derivatives from tessellatus and rubidus. Very shallow soundings, 

 usually less than 40 fathoms, or 240 feet, indicate that Ceralbo and 

 San Jose islands, the home of ceralh&n^is and celeripes, respectively, 

 may have been an integral part of the mainland at a comparatively 

 recent date. 



The large ceralhensis of Ceralbo Island presents more scutella- 

 tional differentiation from tessellatus than any other member of the 

 tessellatus group. Its most striking difference is in the possession 

 of one frontoparietal plate instead of two. It has been shown that 

 the more or less perfect fusion of the normally separated fronto- 

 parietal i^lates into a single piece occurs as an abnormality in tessel- 

 latus, and it is suggested that a population of such variants may 

 have given rise to ceralhensis. Since the pattern of ceralhensis shows 

 a marked likeness to neither that of the geographically adjacent 

 maximus nor the northern ruhidus, one must look elsewhere for a 

 clue as to its possible ancestral stock. While examining specimens, 

 the resemblance of the coarse, dark, longitudinal, dorsal color pat- 

 tern of celeripes of San Jose Island to that of ceralhensis is at once 

 apparent. This suggests the differentiation of ceralhensis and cel- 

 eripes from a common stock. Yet, the derivation of one of these 

 from the other seems unlikely, due to the presence of a depth of 

 about 800 feet between the two respective liabitats as compared with 

 a shallow of less than 100 feet between each and the mainland near 

 it. The presence of red in the ventral and caudal coloration of 

 celerlpes shows its close affinity to mhidus of the neighboring main- 

 land, although the dorsal color pattern is decidedly different, ap- 

 proaching that of northern specimens of tessellatus from the higher 

 levels through a general darkening of the ground color and an em- 

 phasis on the longitudinal in the dorsal pattern. With the change to 

 a darker pattern and the probable adaptation to a brushy, rather 

 than to a sandy habitat, the amount of red in the ventral and caudal 

 coloration has noticeably decreased in celeripes and it is perhaps but 

 secondarily absent in ceralhensis. The influence of the general envi- 

 ronment has been shoAvn, in the discussion of tessellatus (pp. 154-157, 

 IGG), to be great, and the differences between the desert and upland 

 forms here in Lower California seem to be essentially the same as they 

 are in other districts. Therefore, in spite of the superficial distinct- 

 ness of the color pattern of ceralhensis and celeripes from ruhidus, it 

 seems likely that both have been directly derived from that form. 



The four remaining insular forms {canus, martyris, hacatus, and 

 catalinensis) are alike in that they are all dwarf derivatives of tes- 

 sellatus, presenting a similar variation from it — namely, the loss of 

 a longitudinal arrangement in the dorsal markings and the develop- 



