474 TEST ACE A ATLANTIC A. 



stant Cyclostoma} Indeed the G. canariense appears to me 

 to occupy much the same kind of position, in point of variability, 

 as the Clausilia deltostorna does throughout the Madeiran 

 Grroup, — its costal ridges (both as regards their number and 

 development), no less than the intermediate longitudinal sculp- 

 ture, passing through an amount of change which is very ana- 

 logous to what we observe in that protean species, and putting 

 on a different aspect not only for every island but (in a less' 

 degree) for almost every altitude and region in which the shell 

 has become established. 



On this account it is that I have been unable to perceive 

 that the G. adjunctum, Mouss., presents characters of sufficient 

 importance to render its isolation, as a species, either necessary 

 or desirable ; for the peculiarities of sculpture on which it was 

 principally made to rest are so little to be depended upon, and 

 pass into the opposite type by transitions which are so unmis- 

 takeable, that it is impossible, I think, to treat it otherwise 

 than as a variety — as well marked as, but certainly not better 

 defined than, the remaining forms. If however I have under- 

 stood the G. adjunctum aright (and his diagnosis leaves little 

 doubt in my mind upon the subject), Mousson was certainly 

 mistaken in recording it as the phasis which is pre-eminently 

 characteristic of ' Teneriffe ' ; for Grand Canary is the island to 

 which it pertains, and indeed I have not as yet met with a 

 single instance of its occurrence elsewhere throughout the archi- 

 pelago. 



Subtracting the G. Icevigatum from the different aspects of 

 the present species (which I must repeat that I do with a cer- 

 tain amount of reluctance), the G. canariense may be described 

 as a rather large and spirally costate shell, with an intermediate 

 longitudinal sculpture (between the ridges) which varies from 

 minute, closely packed, almost obsolete hair-like lines into com- 

 paratively distant undulating ribs separated by a succession of 

 little pit-shaped impressions,— this latter condition (which 

 occurs in the ' 8- lanzarotensis,'' and which attains its maximum 

 in the ' s. adjitnctus ') causing the entire surface to be decussated, 

 or somewhat reticulate. It has also a great peculiarity about 

 its suture, which overlaps the base of the adjoining volution in 

 the form of a more or less broad and closely-applied lamina, — 

 which latter, however, is more or less irregularly lacerated, it 

 being often deeply gashed like the broken teeth of a saw. This 



' If this surmise should prove to be correct, it follows that the title 

 * Irevigatnm ' (which is the one proposed by Webb for the Gomeran form in 

 1838) will take the precedence over that of 'canansnxc' — which was pub- 

 lished by d'Orbigny in 1839 ; though the latter would, in reality, be far 

 more appropriate. 



