446 



NA TURE 



[April 8, 1922 



narrow vertical strip of black sheet-metal or card- 

 board is fixed across the mouth of the telescope 

 with plasticine, and the telescope is adjusted until 

 the bright image of the slit is completely obstructed. 

 On looking past the strip towards the prisms the 

 observer will see parts of the latter brilliantly lit. 

 By slightly shifting each prism in turn and watching 

 for movement in the bright specks, those due to 

 each prism are readily identified. H. Hartridge. 

 King's College, Cambridge. 



Land Snails of the Madeira Islands. 



In 1892 {Journal of Conchology, vol. vii. No. i) the 

 Rev. R. Boog Watson published a very interesting 

 discussion of the Madeira snails, in which he raised 

 many questions concerning their origin and history, 

 which he did not attempt to answer. To-day, we 

 are still groping for light, but I believe we may reach 

 a number of conclusions which are not likely to prove 

 erroneous. i 



(i.) It is not true that the numerous endemic 

 snails of the islands have come " without trace of 

 descent." They are quite clearly of Palaearctic 

 origin, and their ancestors may be looked for in the 

 rocks of Europe. Among the European Tertiary 

 fossils, Plebecula ramondi Brong., from the Miocene 

 of Germany, resembles the Madeira and Porto Santo 

 Pleistocene fossil P. bowdiichiana ; Pseudocampylaea 

 insignis Klein, from the Miocene of Wiirtemberg, 

 resembles the great P. lowei of Porto Santo, but is 

 not so large ; two species of Craspedopoma, apparently 

 allied to Madeira forms, are found in the Eocene of 

 France. These fossils may be seen in the British 

 Museum. In the absence of anatomical evidence, 

 the relationships of these fossils must remain some- 

 what uncertain, but it is at least probable that their 

 apparent afiinity with the Madeira snails is not wholly 

 deceptive. 



{2.) Considering the diversity and strong peculiarities 

 of most of the Madeira species, it is unlikely that their 

 ancestors arrived later than the beginning of the 

 Tertiary, and it is not improbable that at least part 

 of the immigration dates from the Upper Cretaceous. 

 This postulates a greater age for the islands than our 

 present geological information can confirm. Accord- 

 ing to this view, the European fossils are presumably 

 not the actual ancestors of the Madeira fauna, but 

 derivatives from the same general stock. 



(3.) A comparison between the snails of Madeira 

 and the adjacent island of Porto Santo, which is 

 easily visible from the coast of the larger island, 

 shows that we have two very distinct faunae, with 

 very few species in common. More than this, various 

 genera or subgenera are restricted to one of these 

 islands, or specially characteristic of one. It is, I 

 think, quite certain that during the whole long 

 history of the snail fauna, Madeira and Porto Santo 

 were never united. On the other hand, everything 

 indicates that the three Desertas were during Tertiary 

 time joined together and continuous, or nearly 

 continuous, with Madeira. The following facts are 

 illustrative : 



Pseudocampylaea. Two species, Porto Santo only. 



Cryptaxis. Three species, Madeira and Desertas 

 (Lcptaxis differs^ anatomically, as I shall show else- 

 where) . 



Katostoma. Several forms, Porto Santo and 

 adjacent islets. 



Lampadia. One species, Porto . Santo. (The 

 Madeiran membranacea does not belong here.) 



Idiomela. One species, Porto Santo. 



Hispidella. One species, Madeira. 



Lemniscia. One species, Porto Santo. (I am 



NO. 2736, VOL. 109] 



satisfied that the Madeira calva and galeata are not 

 related.) 



Actinella. Five species, Madeira and Desertas. 



Hystricella. Seven species, Porto Santo and 

 adjacent islets. 



Geomitra. Seven species, Madeira and Desertas. 



This list could be extended, but it is, I think, 

 sufficient to indicate, not only that Madeira and 

 Porto Santo were not united, but that they were 

 never united with the mainland. The diversity of 

 the genera and subgenera on the two islands might 

 be expected as the result of accidental colonisation 

 at very rare intervals, but could scarcely result from 

 the breaking up of an originally homogeneous fauna. 

 If we employ the aggregate genera Leptaxis and 

 Geomitra (properly Ochthephila Beck, which is not 

 preoccupied) without subdivision, the actual facts 

 are obscured. 



(4.) Nevertheless, we have to account for the 

 occurrence of a certain number of identical species 

 in the two islands, and the fact that some of the 

 groups, such as D.scula, are well represented on both 

 islands, with closely allied species. The identical 

 species were not introduced by man, as they occur 

 fossil in the Pleistocene of Cani9al. As a general 

 rule, when a group has representatives on both islands, 

 it appears to be primarily or primitively a Porto Santo 

 group. An apparent exception is Callina, with four 

 Madeira species and one {rotula) in Porto Santo. The 

 species rotula is very peculiar, and probably should 

 not be associated with the others. How these various 

 snails or their ancestors crossed the 23 miles from one 

 island to the other is unknown. Some may have 

 been carried by birds, possibly some may have come 

 on floating pumice or other floating objects.^ The 

 reason for the apparent tendency of Porto Santo 

 types to reach Madeira, rather .than the reverse, 

 may be found in the fact that the arid eastern end 

 of Madeira is well suited to Porto Santo species, 

 while Porto Santo is unsuited to the species from the 

 moister uplands or coasts of the greater part of 

 Madeira. 



(5 ) The occurrence of well-defined species and 

 subspecies on the islets around Porto Santo — some 

 of them no larger than a large building — proves that 

 no important oscillations of level have taken place 

 in recent geological time. Very moderate alterations 

 of level would submerge the islets, or unite them with 

 the main island. The existence of these distinct 

 forms on islets close to the main island also proves 

 that the means of crossing the sea, whatever they are, 

 operate at extremely infrequent intervals. 



(6.) With regard to the species of the Madeiras 

 which are actually identical with those of Europe, 

 it must be said that the presumption is in nearly 

 every case that they were introduced by man. It 

 is possible, however, that some of the smaller ones 

 were brought by " natural " means in geologically 

 recent times, and highly probable that Balea was so 

 brought to Porto Santo, on the feet of birds. Records 

 of the occurrence of European species in the Pleisto- 

 cene deposits of Madeira and Porto Santo all break 

 down on critical examination. 



That the islands are really " oceanic " is indicated 

 by the total absence of indigenous mammals (except 

 bats) and amphibians, and the general character of 

 the invertebrate fauna and of the flora. The multi- 

 tude of snails has seemed to suggest a former land 

 connection, but I now believe that the snails them- 

 selves negative this view. T. D. A. Cockerell. 



University of Colorado, Boulder, March 2. 



1 Experiments should be made to determine whether it is even possible 

 for snails protected by epiphragm or operculum, to pa-<s alive through the 

 alimentary canal of birds. Compare Wallis Kew, "The Dispersal of 



Shells," p. 45. 



