August 5, 1920] 



NATURE 



729 



\ Penck's placing of these pulture stages, which, 

 Mwever, rests only on indirect evidence. 



Taking as his criteria of evolutionary advance four 

 haracters, namely, (i) cephalic index, (2) orbital 

 ulex, (3) hair section, and, in a modified degree, 

 \) skin colour, Dr. Griffith Taylor then attempts an 

 .dialysis of the existing races of mankind, and, so 

 far as the available data permit him, shows that th6 

 more primitive races, or those with low cephalic and 

 virbital indices and relatively flattened hair section 

 (generally associated with depth of skin colouring), 

 have been thrust to the more distant parts, from a 

 niigrant's point of view, of the outlying continents. 

 Closer in to the centre of distribution come races 

 with successively higher indices, rounder, straighter 

 , hair, and reduced colouring, passing through brown, 

 olive, and white to the yellow, brachycephalic, and 

 excessively straight-haired Mongolians, who are the 

 last development of all. So far this is Dr. W. D. 

 Matthew elaborated. A new element is now, how- 

 ever, introduced into the discussion, for an attempt is 

 made to correlate the living with the prehistoric races. 

 One remembers Sollas's tentative comparisons in 

 Ton ("Ancient Hunters"), viz. Tasmanian with 

 Eojithic or Early Palaeolithic, Australian with Mous- 

 terian. Bushman with Aurignacian, and Eskimo with 

 Magdalenian. The author now postulates direct 

 descent for these and many other races too numerous 

 to mention here. 



To^ bring out the nature of the climatic impulse 

 that is supposed to have initiated the successive waves 

 of migration from the homeland in Asia, the author 

 employs the analogv of the seasonal oscillation of the 

 belts of tropical rains, desert, and polar rains in Aus- 

 tralia. As he effectivelv remarks, Nature has placed 

 •Australia like a blackboard, on which are recorded 

 the results of the mobile but very regular and law- 

 abiding climatic zones of the southern hemisphere. 

 He brings forward evidence to show that these zones 

 underwent an analogous migration during the 

 climatic oscillations of the Ice age. 



The analysis outlined above forms parts i.-iii. of 

 the paper. The general exposition of* the argument 

 is crude, and, were it not for the exnUcitness of the 

 diagrams, would be difficult to follow. This tabu- 

 lating and precise definition of material largely 

 speculative gives an illusory impression of the state 

 of the subject, but, if the reader is not misled by it, 

 it certainly conduces to clearness. The adroit 

 handling of a subject so as to distinguish fact and 

 legitimate inference from mere speculation is the last 

 art of the scientific writer. 



The remainder of the paper is devoted to geological 

 speculations of less interest. Part iv. is an exposi- 

 tion of Chamberlin's theory of cyclic change. Cham- 

 berlin's writings (1897-1901) on" the subject are not, 

 however, quoted, the principal authority relied on 

 being Schuchert (1914). In this section a table of 

 roughly approximate temperatures is given for the 

 periods from the Triassic to the present. That this 

 type of tabulated speculation is dangerous is instanced 

 by the fact that both zoological and botanical evidence 

 show that the temperature of Europe in the Neolithic 

 period was several degrees higher than it is at the 

 present day, instead of 7° F. lower, as stated. 



Part V. is an estimate of geological time based on 

 various authorities. The statetnent is made that 

 "Joly quotes similar figures, indicating about 

 300,000,000 as the time interval since the same epoch " 

 (the Cambrian). It is difficult to conceive the author's 

 motive, if any, for this implicit misrepresentation of 

 the works cited, for it matters not a straw to his 

 theory whether the interval since the Cambrian is 

 300,000,000 or less than 100,000,000 years, as concluded 

 by Joly. 



Part vi. is a suggestion, on astronomical lines, of 

 rhythmic oscillations of climate, etc. It is on a par 

 with many former theories of the Ice age in assign- 

 ing a cause which there is no independent evidence 

 to show was ever operative. 



Papers such as this which deal in giddy speculation 

 have for some time past been looked at askance bv 

 the more puritanically minded of our elder geologists. 

 We are not sure that they deserve the contempt 

 with which they are treated. In this matter, 

 however, there is a golden mean, and we should have 

 preferred to see the present paper made less conj- 

 prehensive, and the leading subject-matter of human 

 migrations more thoroughly dealt with. It is of no 

 use trving to straighten out the universe in an article. 



W. B. Wright: 



Long-range Forecasting in Java. 



PUBLICATION No. 5, 1919, of the Royal Ob^^rva- 

 *• tory 01 batavia, entuiea •' Atmospneric Varia- 

 tions 01 jihort and l^ong Duration in the Malay Archi- 

 pelago and Neighbouring Regions, anu the Possibility 

 to Jforecast Ihem," by Dr. k^. braaK, embodies the 

 results 01 a long investigation into the sequence ot 

 raintail in the equatorial regions east of the Indian 

 Ocean. Three kinds of variation are studied : 

 (i) with f)eriods of one or more years up to and 

 including the sun-spot period, (2) secular variations, 

 and (3) with periods less than a month, comparable 

 with Abbot's short-period solar tlucttiations. The 

 variations, the period of which is intermediate between 

 (i) and (3) above, are treated as disturbances of (i). 

 Dr. BraaK lays much stress on a three-year period, 

 of the persistence of which he gives plausible, though 

 not quite convincing, examples. He classifies three 

 groups of years, of high barometer, low barometer, 

 and transition (from high to low), but naturally finds 

 a proportion of years not strictly true to any of 

 these types. It is scarcely surprising that he finds in 

 general a correlation between barometric pressure and 

 rainfall. For the east monsoon he finds strong posi- 

 tive correlation between high pressure and drought, 

 and weaker between low pressure and excess of rain. 

 For the west monsoon he finds, with some local 

 exceptions, excess of rain with high barometer, and 

 deficit with low barometer. His problem is thus reduced 

 to the intensity of the correlation and the chances of 

 a correct forecast of the barometer variation. His 

 next step takes into account temperature changes 

 which may be expected to modify pressure conditions, 

 but his result is disappointing. He obtains rules, but 

 their application is so far a failure that they appear 

 to break down most thoroughly in years of drought — 

 that is, when, if correct, they would' be most valuable. 



Turning to secular variations, he finds no evidence 

 of progressive change in Batavian rainfall; in fact, 

 the only progressive change on which he lays stress is 

 in Batavian air-temperature. Comparison with stations 

 in India, .\ustralia, and other places in the same 

 quarter of the globe provides other types of change, 

 but none agreeing with Batavia, and the question is 

 left unsolved. 



There remain the short-period pressure waves. The 

 equatorial manifestations of these he attributes to a 

 kind of surge, caused by the great disturbances !n 

 higher latitudes, exercising a sucking influence or its 

 converse, with slight variations of the rainfall, less 

 than 10 per cent, of the normal, the effect of which 

 is to compensate the pressure difference by cooling 

 or heating air probably above the 3000-metre level. 



Other variations of rainfall, humidity, and cloudi- 

 ness he considers to be local, and, on the whole, rejects 



