May I, 1919] 



NATURE 



177 



for variability in rate of vulcanisation. In addition, 

 the authors point out that this variability in respect 

 to rate of cure exists in technical mixings with which 

 manufacturers load the rubber. These mixings are 

 largely mineral constitutents in addition to sulphur. 

 It is further contended that the use of strong 

 accelerating agents, such as oxide of lead, tend to 

 obscure the differences produced in raw rubber by 

 the presence of a natural accelerator. 



Recommendations are made to planters which, if 

 adopted, should considerably reduce variability. They 

 are :— 



(i) Dilution of latex to a constant rubber content. 



(2) The use of acetic acid or other weak organic 

 acid (such as formic acid) as a coagulant. 



(3) All coagulating-tanks should be standardised so 

 that the final rubber sheets or crepe are of the same 

 thickness after rolling. 



(4) Conditions of drying and smoking, especially 

 during the early stages, should be kept as uniform as 

 possible. 



It is pointed out that if sheets of rubber are of 

 varying thicknesses the rates of drying will be 

 different, and, consequently, there will be more varia- 

 tion in the biological changes which take place during 

 the early stages of drying rubber. H. W. 



THE SENSITIVENESS OF PHOTO- 

 GRAPHIC PLATES TO X-RAYSA 



A LTHOl'(iH observations have been published on 

 ■^*- the effect of X-rays on photographic plates, the 

 constants of various plates in use do not appear to 

 have been determined. These experiments follow the 

 standard methods of sensitometry of photographic 

 plates to light in respect of exposure of the plate in 

 strips, of development at a standard temperature and 

 for a constant period (namely, hydroquinone at 20° C. 

 for four minutes), and of the subtraction of the density 

 of a fog strip. The density, i.e. the logarithm to the 

 base e of the ratio of the intensity of the incident 

 to that of the transmitted light, was determined by a 

 polarisation photometer. 



The "exposure" E is defined by the relation 

 E = 'V^it/d', where V, volts, is the pressure applied 

 to the Coolidge tube ; it. coulombs, the quantity of 

 electricity passing through the tube during the ex- 

 posure of f seconds; and d, cms., the distance of the 

 focal spot from the photographic plate. This expres- 

 sion gives the energy of the incident rays. Three 

 values for V were used — 31,500, 73,cxxi, and 83,000. 

 The current varied between 003 and 006 milli- 

 ampere, which is lower than the currents ordinarily 

 used in radiography. Experiments are in progress 

 using higher intensities of radiation. 



When the density, D, for a given plate is plotted 

 against the logarithm of the exposure as above defined. 

 a curve similar to those of Hurter and Driffield is 

 obtained. For densities from o to about i the curve- 

 is convex to the logE axis; above that it is straight 

 to densities of 4, the maximum measur<'d. The inter- 

 cept on the logE axis of the straight line produced 

 backwards is the logarithm of the inertia of the 

 plate, which was found to be independent of the 

 development. This result is the same as for exposun' 

 to the visible light. The slope of the straight portion 

 of the curve gives the contrast. .\ high value for the 

 contrast is one of the desirable properties of an X-ray 

 plate. The " speed " of a plate may be tentatively 

 defined as the reciprocal of the exposure required to 

 produce a density of 5. 



' » Abstract of a paper by Mis<; N. C. B. Allen anH Prof. T. H. I.aby read 



before the Royal Society of Victoria on Auen»t 8, 1918. 



The density produci-d in a given plate was found 

 to be constant for a constant value of the exposure 

 yit/d" over the range V 31,500 to 83,000, and for a 

 limited variation of i and t, but not for a large varia- 

 tion of i and t. This means that, for the waver 

 lengths used, the density of a plate depends, not on 

 the wave-length, but only on the energy of the X-rays. 



Plate Inertia Contract Speed 



Paragon ... 074x10' 24 000017 



,, ... i-i8 23 0000096 



Diagnostic ... 0-71 22 000015 



Sunic i-oo 2-35 000012 



Seed I-I2 i-g 0000066 



VVratten 195 22 0000052 



Wellington ... 170 20 0000050 



Imp. X-ray ... 126 i-6 0000036 



Cramer ... 2- 14 19 0000035 



Ilford 219 19 0000033 



Imp. S.R. .. 1-45 1-55 0000028- 



NO. 2583, VOL. 103] 



RAINFALL VARIATIONS. 



A'l" thi- meeting of the Royal Meteorological Society 

 held on .\pril 16, two papers on variations of 

 rainfall were read. The papers are summarised 

 below. 



Mr. .\. .\. Barnes, in his paper on rainfall in Eng- 

 land, the true long-average as deduced from symmetry, 

 stated that it has been usual to assume that the 

 average annual rainfall during any period of thirty- 

 five years can be adopted for obtaining the "long- 

 average" at any rain-gauge, but he considers that 

 the fluctuations which occur between such averages 

 for various thirty-five-year periods tend to show that 

 the basis is somewhat uncertain. By an exhaustive 

 analysis of the annual readings at thirty-eight rain- 

 gauges in England during the sixty-two years 1856- 

 19 1 7, he shows that variations of as much as 5 per 

 cent, on each side of the mean are quite possible 

 when dealing with successive thirty-five-year periods. 

 From these same records it is then shown that far 

 greater consistency in the value of the average can 

 be obtained bv taking periods symmetrical about the 

 end of the yea"- 1886. Both by means of tables and 

 diagrams Mr. Barne.s shows that that date is a very 

 critical one in regard to rainfall in ?3ngland, and that, 

 as a rule, the years before that date were relatively 

 far wetter than years subsequent to it. Hence the 

 balancing of the" earlier wet years by the later dry 

 years establishes the principle of symmetry about that 

 date, and it is shown that by this method the maxi- 

 mum departure from normal which results from 

 faking each of the fifteen long periods .symmetrical to 

 the end of the >ear 1886 does not exceed i per cent, 

 in the case of any of the thirty-eight gauge? which 

 were examined. 



Mr. C. E. P. Brooks's paper was on the secular 

 variation of rainfall. In order to obtain a measure 

 of the secular variation of rainfall during the past 

 thirty to fifty years, correlation coefficients were 

 worked out between the annual rainfall at each station 

 and " timf," the measure of the latter being the 

 number of yoars Ix'fore or after the middle year of 

 the series. This was done for 162 stations distributed 

 over the globe, and the results were charted on a man. 

 This map shows that the (*reater part of the world is 

 divided amonr* a few definite regions of wide extent, 

 in each of which the rainfall has been either increas- 

 in£f or decreasing. The mo.st important area of 

 increasing rainfall is temperate Eurasia (except the 

 western sea-board) ; other areas are south-east .South 

 .\merica and the south of .Australia. .Areas of decrease 

 are the tropical re£?ions as a whole, .South .Africa, and 

 the west coast of EurofX'. It is noted that the number 



