Sept. 4, 1884] 



NA TURE 



447 



be compared with the land-fauna, but whales, porpoises, seals, 

 sirenians, turtles, and sea-snakes are for the most part widely 

 diffused. The best class of the Vertebrata for comparison is that 

 of the fishes, and some details taken by Wallace from Giiniher's 

 "British Museum Catalogue " are very important. The whole 

 class is divided into 116 families, of which 29 are exclusively 

 confined to fresh water, whilst So are typically marine. Of these 

 So no less than 50 are universally, or almost universally, dis- 

 tributed, whilst many others have a very wide range. Four 

 families are confined to the Atlantic and 13 to the Pacific Ocean, 

 whilst a few more are exclusively southern or northern. About 

 63 are found in both the Atlantic and Pacific. 



Now, of the 29 fresh-water families, 15, or more than one-half, 

 are confined each to a single region, 9 are found each in two 

 regions. 2 in three regions, and the same number in four ; one 

 only (Cyprinidte) is found in five regions, whilst not one is met 

 with in all six. It is impossible to conceive a greater contrast : 

 50 marine families, or 62*5 per cent., have a world-wide distri- 

 bution, whilst not a single fresh-water family has an equally 

 extended range, and more than one-half are confined each to a 

 single region. 



The regions adopted by Wallace, as already stated, are founded 

 on the Vertebrata ; he considers, however, that the distribution 

 of the invertebrates is similar. So far as the terrestrial Mollusca 

 are concerned, I am inclined to dissent from this view. But for 

 one circumstance, the Mollusca would afford an admirable test of 

 the theory that marine types — species, genera, and families — are 

 much more widely spread than terrestrial. I am assured that 

 this is the case, but the difficulty of proving it arises from the 

 fact that the classification of pulmonate terrestrial Mollusca, as 

 adopted by naturalists generally, is so artificial as to be worthless. 

 Genera like Helix, Butimus, Achalina, Pupa, Vitrina, as usually 

 adopted, are not real genera, but associations of species united 

 by characters of no systematic importance, and the attempts that 

 have hitherto been made at a natural classification have chiefly 

 beenfounded on the shells, the animals not being sufficiently known 

 for their affinities, in a very large number of cases, to be accu- 

 rately determined. Of late years, however, more attention has 

 been devoted to the soft parts of land mollusks, and in Dr. Paul 

 Fischer's " Manuel de Conchyliologie " now being published, a 

 classification of the Pulmonate Gasteropoda is given, which, 

 although still imperfect for want of additional information, is a 

 great improvement upon any previously available. In this work 

 the first 13 families of the Pulmonata Geophila comprise all the 

 non-operculate land Mollusca, or snails and slugs, and these 

 13 families contain 82 genera thus distributed : — 



Peculiar to one of Wallace's land regions ... 54 

 Found in more than one, but not in both America 



and the Eastern Hemisphere 12 



Common to both hemispheres 16 



The last 16, however, include Umax, Vitrina, Helix, Pupa, 

 Vertigo, and some other genera which certainly need further 

 repartition. The operculated land-shells belonging to a distinct 

 sub-order, or order, and closely allied to the ordinary Proso- 

 branchiate Gasteropoda, are better classified, the shells in their 

 case affording good characters. They comprise four well-marked 

 families [HelieinicUe, Cyclostomidce, Cyclophoridte, and Diplom- 

 matinidte), besides others less well marked or but doubtfully 

 terrestrial. Not one of the families named is generally distributed, 

 and the genera are for the most part restricted to one or two 

 regions. The portion of Dr. Fischer's manual relating to these 

 Mollusca is unpublished, and the latest general account available 

 is that of Pfeiffer, published in 1876 (Monographia Pnemonopo- 

 morum Vivintium, Supp. iii.). From this monograph I take the 

 following details of distribution. The number of genera enu- 

 merated is 64 (including Proserpinida). 



Peculiar to one of Wallace's land regions . . . 4S 

 Found in more than one, but not in both America 



and the Eastern Hemisphere S 



Common to both hemispheres 8 



It is the distribution of the terrestrial operculate Mollusca 

 ■which induces me to suspect that the distribution of land Mollusca 

 differs from that of land vertebrates. One instance I may give. 

 There is nowhere a better-marked limit to two vertebrate faunas 

 than that known as Wallace's line separating the Australian and 

 Oriental regions, and running through the Malay peninsula 

 between Java, Sumatra, and Borneo on the one hand, and Papua 

 with the neighbouring groups on the other. There is in the two 

 regions a very great difference in the vertebrate genera, and a 



considerable replacement of families. The Oriental Vertebrata 

 contain far more genera and families common to Africa than to 

 Australia. Now, the operculate land-shells known from New 

 Guinea and Northern Australia belong to such genera as Cyclo- 

 phorus, Cyclotits, Lcptopoma, Pupinella, Pupina, Diplommalina, 

 and Helicina, all found in the Oriental region, and mostly 

 characteristic of it, whilst the only peculiar types known are 

 LeucoptycAia, closely allied to Leptopoma, from New Guinea, 

 and Heterocyclics, apparently related to the Indian Cyathopoma, 

 from New Caledonia. Farther east, in Polynesia, there are some 

 very remarkable and peculiar types of land-shells, such as 

 AchalineUa, but these do not extend to Australia or Papua. On 

 the other hand, scarcely a single Oriental genus extends to Africa, 

 the terrestrial molluscan fauna of which continent differs far more 

 from that of the Oriental region than the latter does from that of 

 tropical Australia. 



The same is the case with plants. In an important work 

 lately published by Dr. O. Drude of Dresden, the tropics of 

 the Old World are divided into three distinct regions — (1) 

 tropical Africa; (2) the East African islands, Madagascar, &c. ; 

 (3) India, South-Eastern Asia, the Malay Archipelago, Northern 

 Australia, and Polynesia. 



A very large proportion of the families and even of the genera 

 of marine Mollusca are almost of world-wide distribution, and 

 even of the tropical and sub-tropical genera the majority are 

 found in all the warmer seas. I have no recent details for the 

 whole of the marine Mollusca, but a very fair comparison 

 with the data already given for land-shells may be obtained 

 from the first twenty-five families of Prosobranchiate Gastero- 

 poda, all that are hitherto published in Fischer's manual. 

 These twenty-five families include Collides, Olividce, Volutidce, 

 Buccini*i<f, Muriei<Le, Cypreidte, Strombidte, Cerithiidee, Pla- 

 naxidie, and their allies, and contain 116 living marine genera, 

 the known range of which is the following : — 



Found only in the Atlantic Ocean 15 



Found only in the Pacific or Indian Ocean, or 



both 28 



Found only in Arctic or Antarctic Seas or in 



both 12 



—55 

 Found in the warmer parts of all oceans ... 34 

 Widely, and for the most part universally, dis- 

 tributed 27 



—61 

 That is, 52"6 per cent, are found in both hemispheres, whilst 

 only I9"5 per cent, of the inoperculate, and I2'5 per cent, of 

 the operculate land Mollusca, have a similar distribution. This 

 is, however, only an imperfect test of the difference, which is 

 really much greater than these numbers named imply by 

 themselves. 



Some genera of fresh-water Mollusca, as Unio, Anodon, Ci das, 

 Lymnea, Planorbis, Paludina, and Bythinia, are very widely 

 spread, but a much larger number are restricted. Thus, if 

 Unio and Anodon are extensively distributed, all allied fresh- 

 water genera, like Monocondylaa, Mycetopus, Iridina, Spatlia, 

 Castalia, ,'Etheria, and Miilleria, inhabit one or two regions at 

 the most. The same result is not found from taking an equally 

 important group of marine Mollusca, such as Venerida or 

 Cardiada. 



Throughout the marine Invertebrata, so far as I know, the 

 same rule holds good : a few generic types are restricted to 

 particular seas ; the majority are found in suitable habitats 

 throughout a large portion of the globe. The marine provinces 

 that have been hitherto distinguished, as may be seen by re- 

 ferring to those in Woodward's " Manual of the Mollusca," or 

 Forbes and Godwin-Austen's " Natural History of the European 

 Seas," or Fischer's "Manuel de Conchyliologie," or Agassiz' 

 " Revision of the Echini," are founded on specific distinctions, 

 whilst the terrestrial regions are based on generic differences, 

 and often on the presence or absence of even larger groups than 

 genera. 



Botany offers a still more remarkable example. I have just 

 referred to Dr. Oscar Drude's work (Petermann's Mittkeilungen, 

 Erganzungsheft, No. 74, "Die Florenreiche der Erde"), pub- 

 lished within the last few months, on the distribution of plants. 

 Dr. Drude divides the surface of the globe into four groups of 

 floral regions (Florenreichsgruppe), and these again into floral 

 regions (Flor,nreiche), fifteen in number, which are again divided 

 into sub-regions (Gtbiele). The first group of floral regions is 

 the oceanic, comprising all the marine vegetation of the world ; 



