456 



NA TURE 



[Junk 17, igog 



Novcmbir ii, 1907, quoted by Dr. Chrec, stood alone, one 

 would ruadily accept his explanation, but since that date 

 many similar differences, though not quite so large, have 

 been recorded. In (act, the noticeable point about the 

 so-called isothermal layer is the very large differences of 

 temperature that arc found at the same time over places 

 a few hundred miles apart, and over the same place within 

 a [K-riod of twenty-four hours. Ikcause we cannot explain 

 the phenomena, are we, therefore, to doubt their exist- 

 <nre? It is perfectly natural to do so; the question is 

 simply one of the credibility of the evidence. 



The evidence is of various kinds. If one of the instru- 

 ments used in Kngland be completely immersed in a bath 

 of liquid by an observer .^, the temperature of the bath 

 being, say, between +30° C. and —50° C, a second 

 observer H having the record and the instrument can ascer- 

 tain within 1" C, or at the most 2° C, the temperature 

 of the bath used by A. Why, then, cannot B equally 

 well ascertain the temperature of the air through which 

 the balloon has carried the instrument? .Secondly, 

 these instruments are carried up by a balloon travelling 

 through air that has been in contact with the balloon ; 

 the balloon in general bursts, and they fall, moving 

 now .It a much greater speed, since in England we 

 use no parachute. Two traces are made, the one show- 

 ing the temperature during the ascent, the other during 

 the descent, but it is not often possible to say which 

 is which. As a rule, the two traces are quite distinct ; 

 mostly, one indicates a temperature of from 1° C. to 3° C. 

 below the other throughout, but sometimes the traces 

 cross and re-cross each other. However, the point is that 

 the two traces are practically identical ; any peculiarity 

 of gradient shown on the one is reproduced at the same 

 height on the other. Now I think it lies w-ilh those who 

 imply that our instrumental records are untrustworthy to 

 explain this. If the temperatures shown by these two 

 traces are not the approximate temperatures of the air, what 

 .ire they? .Systematic errors could not be the same in the 

 different circumstances of the ascent and descent. It is in- 

 conceivable that casual errors could always so combine as 

 to give errors of the same magnitude in pairs time after 

 time. It is even less likely than that a man, drawing 

 coloured balls from a bag, should draw the same colour 

 in every two consecutive draws, for not only is the general 

 trace reproduced, but every peculiarity in it is also repro- 

 duced. 



Thirdly, the results obtained on the Continent and in 

 .America agree perfectly with those obtained with different 

 instruments and a different system in England. This alone 

 is not a good argument against the possibility of large 

 casual errors, since casual errors are eliminated in the 

 means, but the two sets of observations are as yet not 

 very numerous — about 100 in England — and thev show 

 ihe same general relation between the temperature and 

 height of the isothermal column and the height of the 

 barometer at the surface. 



Or. Three, from the last paragraph of his letter, appears 

 to think that the instrument makers supply the scale. 

 This is not the rase in England, and I do not think it is 

 abroad. .Almost every instrument sent up in England to 

 the present time has been made here. The University of 

 Manchester is responsible for the scales of those th.-it it 

 sends up, and I am responsible for the scales of the rest. 

 These scales are verified before and after e.ich ascent. 

 The lag in our instruments is very small, since we depend 

 on the expansion and ronlrnction of a strip of very thin 

 Oerman silver, but I do not see that the lag affects the 

 general question, since it will be largely eliminated if we 

 take the mean of the ascent and descent. 



\V. II. Dines. 



Pyrlon Hill, Watlinglon. 



.As one who subscribed to the " Confession of Monaco," 

 may I be allowed to say that no definitions of the names 

 stratosphere and isothermal layer were supplied at the 

 conference as those present understood the terms? The 

 meaning of a word has often divided Ihe orthodox from 

 Ihe^helerodox, and for the benefit of I")r. Chree, and also 

 of " heretics in England," I will endeavour to make the 

 rnatter clearer. Rnlloon ascents show that, apart from 

 irregularities near the surface, the temperature of the air 

 NO. 2068, VOL. 80] 



decreases with height fairly regularly up to a certain 

 point ; above this |>oint the regular decrease ceases, and 

 for still greater heights the temperature changes are very 

 small; sometimes there is a small increase, sometimes 

 a small decrease, and sometimes the temperature remains 

 almost constant up to the greatest height reached by the 

 balloon. At any one place and time it thus appears that 

 the atmosphere is divided into two layers, which differ 

 markedly from one another in their vertical temperature 

 distributions. 



A diagram from an actual ascent made here on 

 October i of last year shows the two characteristic 

 temperature gradients. To the upper layer the names 

 isothermal layer and stratosphere have been given ; the 

 latter name is due to .M. Teissercnc de Bort. who surmises 

 that the lower layer, or troposphere, is the part of the 

 atmosphere c o n- 

 cerned in the ver- 

 tical circulation 

 associated with 

 cyclones and anti- 

 cyclones, while the 

 stratosphere lies 

 above such move- 

 ments. The name 

 isothermal layer is 

 not a fortunate 

 one ; certainly none 

 of the orthodox 

 who were assem- 

 bled at Monaco 

 would maintain 

 that the upper 

 layer is isothermal 

 cither in time or in 

 a horizontal direc- 

 tion. Some less 

 misleading term 

 might have in- 

 creased the number 

 of the " elect." 

 Both terms, how- 

 ever, are now in 

 general use, and 

 give definite names 

 to a definite thing, 

 which, as Huxley 

 said, is the object 

 of nomenclature. -m -4o -;o o -jo 



The character- TcMPtn.runt c 



istic temperature- 

 gradient of the upper layer has been found over all parts 

 of Europe, over the .Atlantic, and over North America, 

 but near the equator, if it exists at all, it is at a much 

 higher altitude than in temperate latitudes. Its absence 

 over the equator, and the fact that lower temperatures 

 have been recorded there than in any other part of the 

 atmosphere, seems to me to be a further proof, if such 

 were needed, that the temperature gradient of the upper 

 air recorded in other places is not the result of instru- 

 mental error. Charles J. P. Cave. 



Ditcham Park. Pelersfield, June (>. 



The Sense of Proximity. 



In Nature for March 11 there is an interesting account 

 bv Dr. McKendrick of some investigations by Kunz, of 

 Miilhausen, .uid Prof, liriesbach, on tho senses of the 

 blind. Among other points that he refers to and dis- 

 cusses is Ihe question of the ability of the blind to avoid 

 obstacles and find their way about. This calls to my 

 mind some observations and experiments which I m.ade 

 upon myself some eleven years ago with reference to my 

 ability 10 find my way about with my eyes shut or in the 

 dark. These I had intended to extend and amplify, but 

 up to the present these further experiments have been 

 crowded out by press of other work. 



Many people have the feeling that if, for instance, they 

 are in a room in Ihe dark, they have some perception 

 of their relation to obierls in the room, and particularlv 

 ran appreciate when thev are near one of the walls. \ 

 can remeinber h.iving h.id this feeling for many years, but 

 never had Ihe opportunity of putting it to scientific test 



