April 24, 1903. J 



SCIENCE. 



673 



by Marcliaiid and Fabre ('Les erosions tor- Numerous monadnoek-like islands of conti- 

 rentielles et subaeriennes sur les platens des nental rocks (granite, etc.), often rugged and 

 hautes Pyrenees,' G.-B. duCongres des Soc. mountainous, rise from the shelf as far out 

 8ci. en 1899, Paris, 1900) indicates that the as twenty miles from the mainland. The 

 doubt is well founded and refers the asym- islands and the mainland are commonly bor- 

 metrical form to the action of the northwest- dered with low, sandy coastal plains and man- 

 erly rain-bringing winds and the associated grove swamps, up to twenty miles in breadth, 

 action of lateral rills and radial torrents on exhibiting consequent drainage; and from 

 the weaker and stronger strata of the fans this a slight modern elevation is inferred. 

 A determining factor is found in a compact Some of the islands are made of sand only, 

 clayey layer at about mid-height on the bearing high dunes. Many of the islands are 

 valley side between weaker, sandier deposits tied together or to the mainland by tombolos, 

 below and above. So long as the valleys are bays are more or less enclosed by bars, and 

 worn only in the weak upper layer, their rivers are deflected scores of miles northward 

 cross-section shows a gentle slope on the side by the growth of heavy sand reefs under the 

 AB that is attacked by the northwest winds, action of currents and waves driven by the 

 But when the valleys are worn through the southeast trade wind. The monadnocks in- 

 resistant clays to the weak under layers, the crease in number on the mainland, until the 

 lower slopes, CD, on the wind-attacked side highest part' of the back country gains the ap- 

 pearance of an undulating tableland, up to 



_-_^-_^ ---^^-^^A__ 4,000 feet in altitude. This is described as 



rC-mfimi^^^aw,- showing late-mature Tertiai-y valleys eroded 

 1,000 feet or more beneath a Cretaceous pene- 

 plain, whose remnants are often capped with 



are steepened, although the earlier relation basalt outliers resting on auriferous gravels, 



may still obtain on the higher slopes; and it About Pliocene time the whole country was 



is in this condition that most of the valleys are uplifted so that canons 3,000 feet or more 



found. The explanation of the process by in depth are now cut in the Tertiary valley 



which this change of form is brought about floors; the streams plunge down falls 1,000 



is not immediately convincing and is too de- feet in height .from the as-yet-uncut valley 



tailed for abstract here. floors into the canyon heads. W. M. Davis. 



THE QUEENSLAND COAST. 



A ' Preliminary note on the Geology of the 

 Queensland Coast * * * ' of northeastern 

 Australia, by E. C. Andrews (Proc. Linnean 

 Soc. N. 8. W., 1902, pt. 2, pp. 145-185), pre- 

 sents in modern form — although not in the 

 best arrangement — a highly appreciative ac- 

 count of the mainlaiid and islands back of the 

 Great Barrier reef. The terminology of Gul- 

 liver's essay oh ' Shore line topography ' is 

 largely used. The continental shelf on whose 

 outer edge the great reef is built at from 

 fifteen to one hundred miles from the main- 

 land is described a lowland and platform of 

 subaerial and marine denudation and deposi- 

 tion, moderately submerged in Pleistocene 

 time. The shelf continues south of the reef, 

 its outer slope always rising from great depths. 



RECENT ZOOPALEONTOLOGY. 



COJIPARISON OF THE EUROPEAN AND AMERICAN 

 EOCENE HORSES. 



A PAPER published in March, 1901, which 

 should have been reviewed earlier is by Pro- 

 fessor Charles Deperet, of Lyons, entitled ' Ee- 

 vision des Formes Europeennes de la Famille 

 des Hyracotherides.' It consists of the study 

 and redefinition of all the types of Eocene 

 horses described during the last century in 

 France and England before the ancestral re- 

 lationship -of any of these animals to the 

 horses was appreciated. Since the recognition 

 of the Eocene horses in America by Marsh, it 

 has become evident that they are very closely 

 allied, if not identical in stages of evolution 

 with contemporary forms in Europe. As a 

 result of a close analysis, which is accompan- 



