440 



NATURE 



[December i6, 1915 



ment, the change in density being proportional to the 

 change in temperature. The viscosities were deter- 

 mined in the Redwood I. and II. instruments, from 

 which absolute values can be obtained by means of 

 the relationship jj//3 = AT-B/T, where A and B are 

 calibration constants of the viscometer. The most 

 marked observation in the viscosity results was the 

 relatively enormous temperature coefficient exhibited 

 by the Mexican fuel oils. Further, the results were 

 shown to be dependent on the previous history of the 

 oil. For example, an oil kept at 32° F. for six days 

 showed an increase in viscosity at 60° F. amounting 

 to 20 per cent., while a similar specimen, heated to 93° 

 F. for the same period, exhibited a decrease of 11 per 

 cent. This hysteresis effect was shown to have con- 

 siderable practical importance in the handling and 

 utilisation of such oils. The viscosity of mixtures of 

 Mexican and shale oils was described; the viscosity- 

 concentration curves are not linear, but are sagged, 

 and thus it happens that a relatively small addition 

 of shale oil to Mexican oil causes a considerable 

 decrease in the viscosity. Careful tests were also 

 made respecting the flash points of the oils, and their 

 mixtures. The observation was made that a bulk 

 sample of oil flashed somewhat lower than the small 

 amount used in the Gray instrument. These experi- 

 ments were carried out by Mr. W. F. Higgins. 



Dr. Glazebrook concluded by pointing out that much 

 work remained to -be done in connection with the 

 physics and chemistry of the mixtures of complex 

 bodies which constitute fuel oils. 



QUEENSLAND RAINFALL.^ 



TWrR. H. A. HUNT, the Commonwealth Meteor- 

 -'-*-*■ ologist, has in previous reports given concise 

 histories of the rainfall for New South Wales and 

 Victoria, the volume under notice being thus the 

 third of the series to be published. This report con- 

 tains all the available annual totals and number of 

 days with rainfall recorded to the end of the year 

 19 13 for 1040 stations in Queensland, and in addition 

 monthly values up to the end of 1912 are given for 

 137 stations, so distributed as to afford a good repre- 

 sentation over the area under consideration. Many of 

 the records go back to 1880, and even earlier, so that 

 a good working basis is provided for investigators 

 who require information in regard to local seasonal 

 rainfall, or who may wish to compare the annual 

 variations of rainfall in this part of Australia with 

 those of the other States of the Commonwealth. Every 

 effort was made to obtain trustworthy data, and in 

 addition to the official records, all possible sources o^' 

 information, such as histories of Queensland and old 

 newspaper files, were searched. ~Much labour was 

 Involved in the elimination of discrepancies which so 

 often appear when the same record is published in 

 separate reports. A very useful appendix contains 

 tables showing the monthly and yearly values of the 

 meteorological elements (except wind direction and 

 velocity) at Brisbane from 1887 to 1912. 



The volume is copiously illustrated by maps and 

 diagrams, among which may be mentioned twenty- 

 seven annual rainfall maps covering the oeriod 1887 

 to 1913, a map of mean annual "rainfall, monthly 

 normal rainfall maps, an inierim rainfall map for 

 Papua, and a frost map of Australia. Among the 

 diagrams are several showing the height reached by 

 various floods at the stations belonging to the hydro- 

 metric branch of the service, and graphs giving the 



1 Meteorology of Australia. Commonwealth Bureau of Meteorology. 

 Results of Rainfall Observations made in Queensland, including all available 

 Annual Rainfall Totals from 1040 Stations for all Y^ars of Record up to igi3, 

 together with Maps and Diagrams. Pp. 285. (Melbourne, 1914.) Price 

 10s. 6d. 



NO. 2407, VOL. 96] 



mean annual rainfall of the whole State of Queensland 

 and its sub-divisions. An interesting chronological 

 history of remarkable atmospheric occurrences fills 

 96 pages of the report, the data being discussed back 

 to the middle of last century. It is worthy of note 

 that only four displays of the Aurora Australis are on 

 record, the cases observed occurring In the years 1869, 

 1870, 1894, and 1909. 



For the twenty-six years, 1888 to 1913, the mean 

 annual rainfall for the whole State is 26-50 in., that 

 of 1894, the wettest year, 4039 in., and of 1902, the 

 driest year, 12-63 i^i- The wettest year was thus 52 

 per cent, above the average, while the driest year 

 showed a defect of the same amount. In the fourteen 

 years, 1892 to 1905, only four exceeded the average, 

 the disastrous effect of this long-continued drought 

 being well shown in Diagram A, giving the total 

 number of live stock, which fell from 27 millions in 

 1895 to 10 millions in 1902, the decline being checked 

 in 1903 by a rainfall in excess of the normal. 



The large average rainfall map which accompanies 

 the report shows that the maximum rainfall, indicated 

 by the isohyet of 160 in., is centred in the vicinity 

 of Harvey Creek (lat. 17° S., long. 146° E.), and that 

 for about 80 miles north and south of this point a 

 rainfall exceeding 70 In. falls on the coast and for 

 some distance inland. An annual rainfall of 70° in. is 

 also found on the coast in four small patches located 

 in latitudes 15° S., 21° S., 27° S., and 28° S. The 

 smallest rainfall under 10 in. occurs in the west and 

 south-west of the State south of 23° S. 



It would have materially helped In the elucidation 

 of the problems pertaining to rainfall distribution had 

 the orographical features been indicated on this map. 

 Queensland being in the monsoon region, the heaviest 

 rains occur in summer, while the winter is usually 

 quite dry. 



The appearance in recent years of numerous memoirs 

 dealing with Australian climatology is a marked 

 tribute to the rapidly-growing activity and efficiency 

 of the Commonwealth Service, the example of which 

 in this respect will, we hope, be followed by other 

 weather bureaux In the southern hemisphere. 



R. C. M. 



CHEMISTRY OF FIRE AND 

 EXPLOSIVES. 



IN a recent issue of the Revue Scientifique (Sep- 

 tember 25-October 9) Prof. A. Job has an 

 interesting article upon the chemistry of fire and ex- 

 plosives. After considering the conditions for, and 

 reactions during, ordinary combustion, the connection 

 of explosion with combustion is described. It Is 

 pointed out that, in addition to the volume of gas 

 evolved, heat developed, temperature attained, and 

 pressure, another important factor remains, the 

 rapidity of explosion. This leads to a distinction 

 being drawn between explosives suitable for use as 

 propellents, by reason of their progressive combus- 

 tion, and what are termed in this country "high ex- 

 plosives," where the decomposition is initiated and 

 propagated by shock (detonation). Mercury fulminate, 

 nitrogen iodide, and lead azide are discussed as types 

 of these sensitive detonating explosives. Fulminate 

 Is employed in detonators in France, lead azide being 

 preferred In Germany, It being less sensitive to the 

 decomposing action of moist heat. 



As bursting charges for shells picric acid and 

 trinitrotoluene are discussed. It being pointed out that 

 whilst the former combines with metallic oxides, such 

 as those of Iron and lead, giving very sensitive com- 

 pounds, and hence dangerous, the trinitrotoluene is 

 free from this disadvantage. The higher melting 



