NATURE 



105 



THURSDAY, APRIL 



>9>3- 



DESERT LAND FORMS. 

 Das Gesetz der Wiisteribildung in Gegenwari und 

 Vorseit. By Prof. Johannes Walther. Zweite, 

 neubearbeitete Auflage. Pp. xv + 342; illus- 

 trated. (Leipzig: Quelle and Meyer, 1912.) 

 Price 12 marks. 



DESERT regions have received much atten- 

 tion during recent years, and in this volume 

 Prof. Walther presents a very instructive geo- 

 j graphical study of the north-eastern part of Africa. 

 This is something more than a new edition of that 

 which he published under the same title in 1900, 

 for the fourteen essays on different aspects of 

 desert conditions which there appeared have been 

 recast and rearranged under the headings of the 

 character of the desert, erosion in the desert, and 

 deposition in the desert, together with a fourth 

 chapter in which the evidence for the identification 

 of desert areas in the past history of the earth is 

 assembled. A visit to Egypt and the north Sudan 

 in 191 1 provided the opportunity for extending and 

 supplementing his earlier observations, and recent 

 railway extensions enabled him in the time at his 

 disposal to visit the oasis of Kharga, Khartum, 

 land to cross the Red Sea hills between the Sudan 

 plains and Suakin. The result is a very interesting 

 and instructive work dealing primarily with a part 

 of the north African desert, but introducing many 

 examples from other arid regions of the world. 



In treating of precipitation in the desert the 

 author has, we think, in following- Sickenberger, 

 gone somewhat too far in saying that dew is 

 entirely absent in the interior of the desert. The 

 absolute humidity is usually not very low, and on 

 cold clear nights dew is not infrequently formed. 

 Stress is rightly laid on the action of rainfall in 

 areas which are fairly described in general terms 

 as rainless, for such falls of rain are not at all 

 uncommonly reported when a wide region is con- 

 sidered, though each fall may be extremely local 

 in extent. The rock tombs on the west of the 

 Nile at Thebes are quoted as providing evidence 

 that the water from the Nile does not there per- 

 colate to any distance from its bed. Here the 

 river flows in its alluvial flood plain for the most 

 part. Grabham has shown that the varying levels 

 of the Blue Nile are to be traced so far as 

 900 metres from its banks in the Sudan, and it 

 appears from discharges which have been measured 

 that a considerable loss from the river takes place, 

 over and above that due to evaporation, in such 

 long reaches as that near Dongola, where the river 

 flows for long distances in the Nubian sandstone. 

 NO. 2 2 66, VOL. 91] 



An especially interesting part of this book deals 

 with the Sudan desert and the "half-desert" on 

 the northern fringe of the monsoon rains, where 

 the extreme aridity of the Nubian and Libyan 

 deserts gives place to less inhospitable conditions 

 and vegetation can develop to a limited extent. 

 The Red Sea hills furnish most instructive 

 instances of this, and of the erosion charac- 

 teristic of such regions. Many interesting 

 examples of erosion and of deposition are 

 described, and are particularly well illustrated by 

 characteristic and well-chosen photographs. 

 Doubtless because the parts of the country 

 which the author visited do not exhibit good 

 examples of the process, there is not much refer- 

 ence to the filling of wide valleys and depressions 

 with rock waste swept down from the higher levels 

 which may be seen so finely represented further 

 north than the Berber-Suakin route which was 

 followed, in the valleys of the complex of crystal- 

 line ranges which form the western shore of the 

 Red Sea. This greatly enlarged edition of a work 

 already well known will be most acceptable to 

 both geographers and geologists. H. G. L. 



THE PROPERTIES OF STEAM. 

 The New Steam Tables: together with their De- 

 rivation and Application. By Prof. C. A. M. 

 Smith and A. G. Warren. With an introduction 

 by Sir J. Alfred Ewing, K.C.B., F.R.S. Pp. 

 xii+101. (London: Constable and Co., Ltd., 

 191 3.) Price 45. net. 



PROF. CALLENDAR, in his Royal Society 

 paper of 1900, suggested the use as the 

 characteristic for steam of v — b = R0jp — CB- n = 

 V-c, say. This is suggested by the Joule- 

 Thomson equation for gases, where 11 = 2, and 

 by Grindley's result for steam, in which 

 11 = 3-8. Only a man of Prof. Calendar's re- 

 putation could have received attention, for 

 he gave rather fanciful reasons for taking 

 n = 3"5, and for his values of the specific heats 

 when p is very small. Again, it is probably quite 

 untrue that c is a function of temperature only. 

 Nevertheless, when steam tables are calculated 

 by means of the above characteristic, the con- 

 slants b, C, and 1? (and, indeed, R also) can 

 be given such values as make the calculations 

 agree with what Prof. Callendar regards as 

 the best experimental results, and he recom- 

 mended in 1900 that tables calculated from his 

 formulae should be substituted for the usual tables 

 as given by Regnault and modified by Griffiths 

 and others. The numbers of the new tables are 

 consistent with each other, and this is a great 

 advantage, because we generally need differences 



F 



