September 9, 19 15] 



NATURE 



53 



had to devote himself to military duties, Dr. Arthur 

 Smith was appointed acting-professor of anatomy in 

 the University of Sydney, and Colonel Wilson most 



» magnanimously handed over to him this most import- 

 ant human relic for investigation and description. 

 With the valuable assistance of Mr. D. M. S. Watson 

 he has succeeded in removing the matrix and fullj 

 pxposing the extraordinary features of Australia's 

 earliest known human inhabitant, perhaps a member 

 of the earliest family to ferry across Wallace's line 

 and tramp dogs through Papua to Queensland. 



The skull has been purchased by the Hon. Joynton 

 Smith, M.L.C., and presented by him to the Uni- 

 versity of Sydney. The formal presentation was made 

 the occasion of a public meeting in the University on 

 June 18, presided over by the chancellor, when Acting- 

 Prof. Arthur Smith and Prof. David delivered ad- 

 esses on the significance of the skull and its 

 laring upon the problems of man's antiquity in 

 .ustralia. 



Prof. David explained that the skull (of the Pleisto- 

 cene period) was found after a flood thirty-one years 

 ago in the bed of a creek near Talgai station, on the 

 Darling Downs, by a stockman. From its highly 

 mineralised state and mode of occurrence in a region 

 where other fossil remains had also been found, the 

 skull was of sufficiently striking geological antiquity 

 to be grouped with the last ice epoch of the northern 

 hemisphere. At the close of the last glacial period, 

 gigantic animals became extinct, and probably the big 

 animals known to have existed in Australia ages ago 

 likewise died at the same time as those of the northern 

 hemisphere. The discovery of the skull might serve 

 to explain how it was that Australia, with its mar- 

 supial fauna, had an invader in the form of the dingo, 

 which was a sort of Asiatic jackal, and "had no more 

 right here than the Germans had in Belgium." There 

 was the possibility of its having been brought here by 

 early man. However, the skull was worth its weight 

 in gold. The University of Sydney and men of science 

 throughout the world were under a deep debt of grati- 

 tude to Mr. Smith for purchasing the skull, which was 

 undoubtedly a priceless specimen. 



Acting-Prof. Smith stated that the condition of the 

 teeth was such as to show that the skull was a youth's, 

 while at the same time it had many of the character- 

 istics of an adult skull. The canine tooth of the 

 Talgai fossil was the largest human tooth so far dis- 

 covered. The extremely primitive characteristics were 

 such as to warrant one claiming for the skull con- 

 siderable importance in anthropology, and worthy of 

 being placed with such specimens as the prehistoric 

 Heidelberg jaw and the Piltdown skull. There was, 

 he concluded, certain to be a great controversy con- 

 cerning the latest discovery when the results of the 

 investigation were made known. 



THE RAINFALL OF JAVA. 

 T YING as it does midway between India and 

 ■'-' Australia, Java forms a climatological link 

 between the two great British Dominions, and we 

 recognise the appropriateness of the compliment 

 implied in the choice of English as the second 

 language in which this important outcome of Dutch 

 enterprise is published. Tables are given of the 

 mean monthly rainfall, number of rain days, and 

 mean daily maximum rainfall (for each month) at 

 106 1 stations in Java from observations taken between 



1 " Uitkomsten der Regenwaarnemingen op Java." (Remits of Rainfall 

 Observations in Java). By Dr. W. van Bemmelen. Pp. xxiii+173. 

 Kegenatlas (Rainfall-Atlas), vii maps. (Batavia : Javasche Boekhandel 

 Urukkerij, 1914 and 1915.) 



NO. 2393, VOL. 96] 



1879 ^Jid 191 1. Dr. van Bemmelen discusses the 

 trustworthiness of the data, which are far from uni- 

 form. Obviously fallacious records have been 

 eliminated, and there is reason to believe ■ that the 

 data here given compare well with those available 

 for other tropical countries. Observations were rtiade 

 to the nearest millimetre, so that the critical rainfall 

 which determines whether or not a day is to be 

 counted as a rain day is 05 mm., or 002 in., i.e. 

 four times the amount taken as determining a rain 

 day in this country. As an example of the variability 

 of monthly rainfall in Java, the case is cited of one 

 station where in September, 1902, the rainfall was 

 12 mm., and in September, 1904, 896 mm. The 

 greatest average annual rainfall recorded at any 

 station in Java is 8305 mm., and the least 882 mm., 

 but for both stations quoted the average was only for 

 the short period of seven years, and experience shows 

 that runs of seven years may occur with every year 

 above or every year below the long-period average. 

 The wettest year ever recorded at any station was 

 10,112 mm. at Sirah Kentjong, Kediri, in 1909, and 

 the greatest fall in twenty-four hours was 511 mm. at 

 Besokor, Semarang, on January 31, 1901. Perhaps 

 I may be pardoned if I remind those who keep 

 metric and vulgar statistics in separate compartments 

 of their brains that these figures are equivalent to 

 398- 1 and 201 in. respectively. 



The most important part of Dr. van Bemmelen 's 

 work is his series of maps of average rainfall. That 

 he was fully alive to the difficulties in his way is 

 made plain by his quotation of the reasons which 

 deterred h?m in 1908 from making the attempt. To 

 our mind the greatest difficulty, and one that has 

 not been overcome, is the use of averages of vary- 

 ing lengths without reduction to a common period. 

 We agree with the author that such reduction was 

 difficult, if not impossible, in the circumstances, and 

 that any sort of isohyetal map is better than none, 

 and for many purposes, both scientific and economic, 

 the maps he has produced will prove most useful. 

 He gives a large-scale (i : 1,000,000) map showing the 

 position of the stations, and a series of smaller maps 

 (scale I : 1,500,000) for annual rainfall, and the mean 

 monthlv rainfall of four selected seasons. These 

 are : — (i) the mean of the three wettest months, 

 December, January, February, when the west mon- 

 soon is blowing ; (2) the mean for the two months 

 March and .April, showing the main transition ; and 

 (3) the mean for the two months May and June, when 

 the transition is almost complete, to (4) the period 

 of the east monsoon, represented by the mean of the 

 three months July, August, September — the driest 

 part of the year ; finally, (5) the mean of October and 

 November, showing the transition to the wet season 

 again. These maps are admirably executed and most 

 suggestive; but we are of opinion that both the 

 climatic transitions and the geographical distribution 

 would be more clearly suggested by maps based on 

 the homogeneous average of, say, the last ten years, 

 than on the data actually used. We are in hearty 

 agreement with Dr. van Bemmelen 's method of 

 drawing isohyetal lines from such data as are avail- 

 able by what he aptly terms "a continual compromise 

 between deduction and induction," the deduction 

 being based on the recognised relationship between 

 configuration, wind-direction, and rainfall, the induc- 

 tion being from the data which have been actually 

 observed. Even where the data are as numerous, 

 long-continued, and reasonably accurate as they are 

 for the British Isles, the most probable distribution 

 of rainfall can only be represented cartographically 

 by such a "continual compromise." 



Hugh Robert Mill. 



