SUMMARY. 5G5 



further to the south) to possess a fauna in ivhich the ' Medi- 

 terranean'' element is much more traceable. This latter cir- 

 cumstance, which is shadowed forth likewise by the Coleop- 

 terous statistics, is by no means a fanciful one, — whole groups 

 which are indicative (more or less) of Mediterranean countries, 

 but which have no single representative elsewhere in these sub- 

 African archipelagos, being quite at home at the Canaries. 

 Thus the section Hemicycla of the genus Helix, which does not 

 even put in an appearance at the Azores, Madeiras, and Cape 

 Verdes, has no less than 37 exponents (indeed probably more) 

 at the Canaries; and the same might be said of the section 

 Turricula, Beck, which is so strongly developed on the opposite 

 coast of Morocco. Cyclostoma, too (as distinguished from 

 Craspedopoma), is dominant in nearly all the Canarian islands, 

 while totally absent from the other Groups ; and it is only in 

 the Canarian cluster that the Mediterranean genera Parmacella 

 and Leucochroa have been brought to light. Moreover several 

 of the European types, of a submaritime habit and a widely 

 acquired range, which occur equally in the other archipelagos, 

 and which I had eliminated from the general catalogue (as with- 

 out signification) when discussing the purely Atlantic element 

 in the faunas, seem to possess a significance at the Canaries 

 which they can hardly be said to do in the neighbouring clus- 

 ters, — having the appearance there of being positively in- 

 digenous, rather than naturalized. Thus, unless I am greatly 

 mistaken, the Helix lenticula, Fer., and the Stenogyra de- 

 collata, Linn., are found in Fuerteventura and Grand Canary in 

 a genuinely subfossilized condition, as are also the latter and the 

 Helix impugnata, Mouss. (which is scarcely more than an ex- 

 treme development of the H. pisana), in Lanzarote. Facts like 

 these render it at least probable that the particular forms to 

 which I am alluding, and which are usually defined as ' Medi- 

 terranean ' ones, have not been introduced into at any rate the 

 Camarian islands since the occupation of the latter by man, and 

 indeed that their presence there is due to natural causes, opera- 

 ting at a remote epoch, — such, for instance, as a slow system of 

 migration over a continuous land, stretching in a north-easterly 

 direction along what is now the coast of Morocco. At all events 

 some such connective tract would answer the requirements of 

 the present fauna, and solve many a problem which it is other- 

 wise difficult to interpret. 



How this theory may be brought to bear upon the principle 

 of segregation, as now witnessed in so many of the archipelagos, 

 it might perhaps be worth while to pause for a few moments to 

 enquire ; for it seems to me that if a more or less continuous 

 land, which may be supposed to have occupied this particular 



