March 27, 1896.] 



SCIENCE. 



473 



books according to the approved plan for 

 local use. The latter suggestion is one that 

 may be commended to the councils of our 

 American and National Geographical So- 

 cieties. 



THE DANUBE. 



A COMPENDIOUS volume on the Danube, 

 by Schweiger-Lerchenfeld (Die Donau als 

 Volkerweg, Schiffahrtstrasse und Eeise- 

 route, Vienna, Hartleben, 1896, 950 p, with 

 many and excellent illustrations and maps) 

 contains much material for the physiog- 

 rapher ; truly not the result of original in- 

 vestigation now first published, but well 

 summarized from many sources and ac- 

 ceptable for those who have to study this 

 great international river at a distance. 

 Most serviceable is the description of the 

 various features of the great Hungarian 

 plain, the Alfold, as it is locally called, 

 through which the Danube and its chief 

 tributary, the Theiss, wind their courses. 

 Sand dunes make deserts of large areas ; 

 other parts are wet and marshy beyond re- 

 demption, and a third division includes the 

 Puszta, or fertile grassy plains. Many dis- 

 tricts have been subject to overflow ; but 

 these are now reduced by the ' regulation ' 

 of the larger rivers, as well as by the con- 

 struction of dikes. Below the ' Iron gate ' 

 in the Carpathians, the course of the 

 Danube has been changed at several points 

 by sand blown into its channel by the south- 

 east storm wind, the ' Koschava,' from an 

 extensive area of ridged dunes. The various 

 narrows of the great river and their improve- 

 ment for navigation are fully described. 



THE LOCATION OF SETTLEMENTS. 



Dr. a. Hettner, Privatdocent in geog- 

 raphy in the University of Leipzig and 

 editor of a new geographical journal, con- 

 tributes to it an essay on the geographical 

 controls of human settlements, reviewing 

 the previous literature of the subject and 

 laying down lines along which further re- 



search should be conducted (Hettner's 

 Geogr. Zeitschr., i., 1895, 361-375). Some- 

 what as plants and animals are affected in 

 their distribution by geographical environ- 

 ment, so man himself responds to his sur- 

 roundings ; his personal will having a much 

 less influence than would appear at flrst 

 sight, although complicating the reaction in 

 a manner not apparent in the case of lower 

 organisms. Just as the features of the land 

 are now best explained by an appropriate 

 historical method of study, based on their 

 geological evolution, so the location of set- 

 tlements should be studied in relation to 

 their development from their beginnings, and 

 not only in relation to their actual sur- 

 roundings. The article as a whole is an 

 abstract consideration of the subject, with- 

 out illustration by specific examples. 



MIDDENDOEFF's PERU. 



A RESIDENCE of twonty-fivc years in Peru 

 affords Middendorfi^ an extended experience 

 for record in his work on that country, of 

 which the third volume, Das Hochland 

 von Peru (Berlin, 1895), now follows the 

 second, Das Kiistenland (1894). The 

 coastal desert belt, with its irrigable val- 

 leys, rises into the highland through dull 

 slopes of rock waste, seldom varied with 

 ledge or cliff, but sometimes trenched by 

 great ravines. Ascending this western 

 slope, the traveler finds himself on lofty bar- 

 ren plateaus, of rather cool climate, holding 

 lakes in their depressions; a special account 

 being given of Titicaca and its surroundings. 

 Very different from the barren ravines of 

 the dry western slope are the deep warm 

 valleys of the rainy and forested eastern 

 slope, in which many streams that head 

 west of the eastern range cut their path on 

 the way to the Amazon. 



As in so many books of travel, this one, 

 although the work of an interested ob- 

 server, loses greatly in geographical value 

 from an insuflicient understanding of physi- 



