February 20, 1896] 



NATURE 



Z17 





sents zenithal distance of a fixed sun, the intensity of which 

 would be 3/10 of the true intensity, and the annual effect of 

 which would be equivalent to that of the true sun. At sea- 

 level this quantity of solar heat is supposed to reduce itself, 

 according to Bouguer's law, from I \.o p""^'^, where/ is the mean 

 transparence of the atmosphere. 



Upon these assumptions, the annual solar temperature of the 

 air on a continental globe is expressed (omitting here a little 



term for diffused heat) by a formula /, = /<^ + k, cos C^ where 



nis is the transparence of the atmosphere for the earth's 

 radiations, and k^ is a coefficient projwrtional to the heating 

 fower of solar radiation for soil, and depends then upon the 

 physical constitution of the latter. On the sea the effect of solar 

 heat is most complicated, but, abstracting from currents, argu- 

 ments are given for accepting on oceanic globe a formula 



tg = tc + i'a ^ — , where nia, k,, are the. analogues of nis, ks for 



sea. Forbes expresses the mean temperature of each parallel by 

 / = ^0 -f Jc (/, - Z^), when x is the fraction of parallel occu- 

 pied by continents ; but the same formula may express the 

 mean temperature of every point on continents, if x 

 means its continentahty, whose expression by annual range e 



■ ^* I is discussed 



The calculation of coefficients in the complex formulae so ob- 

 tained is very much simplified by comparison with known 

 empirical formula;. Mendeleefs formula for vertical distribu- 

 tion of temperature leads to the fundamental fact that the mean 

 temperature of the sky, tc, is a constant for all points at sea- 

 level, and from evaluations of Mendeleefs constant by Woeikof 

 I assume it as -45""4C. (-49° 7 F. ). So is numerically ex- 

 pressed what Mr. Culverwell calls the blanketing function of 

 atmosphere, and its function of conforming temperature on the 

 earth's surface ; and CroU's fallacy of evaluating effects of solar 

 heat by differences from a hypothetical temperature of space 

 is placed in evidence. Values of 4, t^ at equator, discussed from 



Forbes' formula, give values of -, — ' ; / is assumed as 0"6o. 

 in, via 



Solar temperatures so obtained give a very satisfactory repre- 

 sentation of facts ; their differences from true temperatures are 

 a striking reflex of the distribution of meteorological and physical 

 agents (sea and air currents, convective motions, shore-ice, &c.) 

 not accounted for in deduction of formula?. 



An attempt is made for a theorj' of annual range, where 

 Wilson's principle of the constancy of nocturnal cooling of 

 bodies, whatever may be the temperature of the air, allows the 

 assumption that the temperature of the sky follows in its 

 variations temperature of soil. Comparison of theory with facts 

 indicates the enormous smothering influence of meteorological 

 agents. 



Discussions of astronomical and geographical theory of 

 an Ice Age, in the light of the formulae given, accounts for vari- 

 ability of physical and meteorological agents. None seems to 

 me to satisfy either the theory or the actual conditions for 

 variation of climates, as developed on facts by Bruckner's 

 classical work on " Klimaschwankungen." These conditions 

 seem to the author to be satisfied by the suggestion of a small 

 ditninution in the transparence/, attended by a proportional, or 

 by a smaller, diminution of m^, ma- So the difference t - tc is 

 diminished, less at low latitudes, more at higher, above all at 

 70° Lat. , the variation diminishing further. Diminution is 

 greater on sea and less on continent, so diminishing at high 

 latitudes thermic difference between earth and sea, what is, 

 according to Bruckner's demonstration, the capital condition 

 for a rainy j>eriod on continents, and indirectly for glaciers push- 

 ing forward. Also the difference between equator and poles is 

 strengthened. 



Inversely, an increase of / and m^ in„ would bring, as in 

 Tertiary periods, a more equal distribution of temperature between 

 equator and poles, by increasing for several degrees the tem- 

 perature in higher latitudes. Mars is probably in a similar 

 condition, as polar ice-caps do dissolve, notwithstanding that the 

 intensity of sun is there much less than on earth ; but the Martian 

 atmosphere is extraordinarily more transparent than ours. 

 Annual range would be strengthened ; but the flora of East 

 Siberia suggests that also Heer's polar floras might have sustained 

 severe winter, provided that summer's heat was sufficient to 

 support them, and that abrupt variations were avoided. 



NO. 1373. VOL. 53] 



THE RONTGEN RAYS. 



'T'HE field of investigation opened up by Prof. Rontgen's 

 researches on the new actinic rays has attracted explorers 

 from all parts of the civilised world. So numerous are the 

 communicatious being made to scientific societies that it is diffi- 

 cult to keep pace with them, and the limits of our space would 

 be exceeded if we attempted to describe the whole of the 

 contributions to the subject, even at this early stage. It may 

 assist, however, in the organisation of the facts if we bring 

 together a few of the results obtained since the publication of 

 Prof. Rontgen's paper. 



The most important British communication on the subject was 

 made by Prof. J. J. Thomson to the Royal Society on Thursday 

 last, in the following paper, on the " Discharge of Electricity 

 produced by the Rcintgen Rays, and the effects produced by 

 these Rays on Dielectrics through which they pass. 



"The Rontgen rays, when they fall upon electrified bodies, 

 rapidly discharge the electrification, whether this be positive or 

 negative. The arrangement I have used to investigate this 

 effect is as follows : The Ruhmkorff coil and the exhausted 

 bulb, used to produce the rays, are placed inside a large packing 

 case covered with tin plate ; this is done to screen off from the 

 electrometer any electrostatic disturbance due to the action of 

 the coil. The needle of the electrometer is suspended by 

 a quartz fibre ; thus, as there is no magnetic control, the needle 

 of the electrometer is not affected by changes in the magnetisa- 

 tion of the core of the coil. 



'"The exhausted bulb is placed so that the phosphorescent 

 part of it is about li inches from the top of the box, and a hole 

 about an inch in diameter is cut in the lid of the box just over 

 the bulb, so as to allow the rays to emerge from the box ; a 

 thin plate of either aluminium or tinfoil is used to cover up the 

 hole. The electrified plate, which is a little larger than the 

 hole, is placed outside the box about two inches above the hole 

 in the lid, so that the Rontgen rays which passed through the 

 hole fall upon the plate. This plate is kept permanently con- 

 nected with one of the quadrants of a quadrant electrometer ; 

 the greatest care is taken with the insulation of this plate and of 

 the quadrants of the electrometer. The insulation was so good 

 that there was no appreciable leak when the coil was not in 

 action. The following is the method of making the experi- 

 ments : The two pairs of quadrants are connected together, 

 and the plate charged to a high potential by an electro- 

 phorus, or by temporary connection with a large battery 

 of small storage cells. All the quadrants of the ' electro- 

 meter are now at the same potential. The two pairs 

 of quadrants are now disconnected ; if the insulation is 

 good the potentials will remain the same, and there will be no 

 deflection of the electrometer ; in our experiments the leak is so 

 small that under these circumstances the movement of the spot 

 of light is hardly perceptible. If, now, the Rontgen rays are 

 directed on to the plate a violent leakage of electricity from the 

 plate occurs, the potential of the quadrants connected with the 

 plate changes, and in a few seconds the spot of light reflected 

 from the mirror of the electrometer is driven off the scale. This 

 leakage of electricity occurs whether the plate is positively or 

 negatively electrified ; if the plate is uncharged to begin with, I 

 have not been able to detect that any charge is acquired by the 

 plate by exposure to these rays. When the potential to which 

 the plate is raised is high the leakage from the plate is a most 

 delicate means of detecting these rays, more so than any photo- 

 graphic plate known to me. I have found these rays produce 

 distinctly perceptible effects on a charged plate after passing 

 through a zinc plate a quarter of an inch thick. The charged 

 plate and electrometer are much more expeditious than the 

 photographic plate and more easily adapted to quantitative 

 measurements. 



" To determine how the radiation of the Rontgen rays depended 

 upon the degree of exhaustion of the bulb, the bulb was kept in 

 connection with the pump and the leakage was observed at 

 different degrees of exhaustion ; no leakage could l^e detected 

 until the pressure was so low that phosphorescent patches ap- 

 peared on the bulb, and. even after the phosphorescence 

 appeared, the leakage was small as long as there was any 

 considerable luminosity in the positive column ; it was not unt:l 

 this had disappeared that the leakage from the charged plate 

 became rapid. 



" If the greatest sensitiveness is required, it is, of course, advis- 

 able to charge the plate as highly as possible. The leakage due ti> 



