FORMER LAND CONNECTIONS. 323 



pterygians. One portion of the argument is, however, doubtful. 

 The authors suppose that Madagascar has been separated from 

 Africa ever since the middle of the Secondary period, which, if 

 true, would certainly imply a surprising antiquity for the 

 (Jichlidae, as a purely freshwater group. However, the presence 

 of Cichlid perches in Madagascar can be well reconciled with a 

 non-marine origin for all the modem genera, and with their 

 specialised structure, clearly pointing to a fairly recent origin, on 

 the hypothesis of a temporary Miocene connection between Africa 

 and Madagascar, which is considered probable on geological 

 ground's. The Cichlid fauna of Madagascar is actually very 

 scanty, comprising only five species, whereas 288 species are 

 recorded from Africa. This seems to favour direct spreading from 

 Africa rather than an independent marine origin, especially as 

 two of the three Malagasy genera are very closely related to 

 Congo genera. Eegan suggests that Madagascar " may have 

 received its Cichlidae from Africa at a time when it was only 

 narrowly separated from or even temporarily connected with that 

 continent, and perhaps from India when the islands of the Indian 

 Ocean were more extensive and a brackish-water fish might pass 

 from one to another: this time can hardly have been later than 

 the beginning of the Miocene." 



(F.) There remains for brief consideration the remarkable 

 paper on " Climate and Evolution " by Dr. W. D. Matthew. ^^ 

 The author approaches the subject as a firm believer in the 

 principle " of the permanence of the great ocean basins. This 

 depends on the known facts in regard to Isotasy : the rocks under- 

 lying the oceans are heavier than those underlying the continents. 

 Yet, in a footnote he adds the following: — " The suggestion of 

 Bailey Willis that the present Isostatic compensation may be 

 unusually complete must be borne in mind." Dr. Matthew will 

 only admit that certain minor changes in the relations of land 

 and sea have taken place. He cannot believe that Madagascar 

 has had any direct land connection with Africa, for its fauna, 

 he believes, might easily have been obtained through such 

 accidental agencies as floating rafts. This agency is even exalted 

 to explain the case of the South American Hystricomorph rodents, 

 thus: " Oversea transportation froin Africa appears to be the 

 only reasonable interpretation of the evidence at hand." To 

 many of us, such extensive transportation seems inevitable, and 

 not even sufficient as an explanation of the Malagasy fauna. 

 Besides Cichlid perches, which can tolerate brackish conditions, 

 but are quite unknown in the sea, the fauna of Madagascar 

 includes such freshwater animals as crayfish, frogs, tortoises and 

 snakes, all related to freshwater animals of other regions, but 

 intolerant of marine conditions. Not tO' mention the various 

 mammals and land-tortoises, there are ground-dwelling frogs, like 

 Rana labrosa, which seem ill-adapted for transportation on or in 

 an ocean raft. Nor should it be assumed, as Dr. Matthew does, 

 that the recent Aepyornis, a typical ratite bird, arrived on the 

 island by flight, for the evidence that ostriches and their kin could 

 ever fly is not at all satisfactory. 



