282 



NATURE 



[January 7, 1909 



has been met would, on the above supposition, receive a 

 simple explanation in the varying mean densities of the 

 balloon and its contained gas on different occasions, since 

 the balloon will come to rest when its mean density is 

 equal to that of the air in which it floats. 



In this connection it may not be out of place to recall 

 that an increased velocity of the balloon does not indicate 

 an ascensional current, nor does a decreased velocity 

 indicate a downpour of air. In the latter case, what has 

 more probably happened is that the balloon has crossed 

 a surface of discontinuity in density, and is less buoyant. 



J. I. Craig. 



Survey Department, Giza, Egypt, December s, 1908. 



It does not appear to nie that the explanation suggested 

 by Mr. Craig is tenable. I will first state that the instru- 

 ments used in England are calibrated over the whole range 

 of conditions to which they may be exposed by placing 

 them in a glass vessel so that they are completely covered 

 by liquid which has been cooled to the desired temperature 

 by solid COj, and then exhausting the air by a pump. 

 Thus the instrument is exposed at the same time to the 

 conditions of pressure and temperature which it will meet 

 with in use. This is done both before and after each 

 ascent, unless, as sometimes happens, such damage is done 

 by the finder as to render the second calibration impossible. 

 On the Continent, at one station at least, and perhaps at 

 most, the pressure is reduced slowly for the express pur- 

 pose of meeting the point raised by Mr. Craig. In 

 England, and for the same reason, air is generally left in 

 the aneroid box. Very thin metal is used ; the box is dried, 

 the faces are squeezed together so that they nearly touch, 

 and the box is then sealed up. The result is that the 

 pressure scale depends on the elasticity of the enclosed air 

 chiefly, and only slightly on the elasticity of the metal. 

 Of course, there is a large correction for temperature 

 which involves extra trouble in the calibration, but, on the 

 whole, I believe this system to be the more accurate. 



The lag of an aneroid box no doubt produces the results 

 described by Mr. Craig, but the error so produced, when 

 expressed as a percentage of the whole deflection of the 

 box, is, I think, small. We have many records in which 

 the isothermal part of the trace covers more than 100 mm. 

 of pressure, and an error of this magnitude could not 

 possibly be produced by the elastic fatigue of the box. 

 Further, in general the balloons burst while they are 

 rising ; they start with an ascensional velocity of more 

 than 600 feet per minute, and we have good ground for 

 thinking that the pace is accelerated towards the top. 

 They fall in about half the time they take to rise, and 

 hence, unless there be an isothermal layer, the instrument 

 cannot remain in air at the same temperature for sufficient 

 time for the slow expansion of the bo.x to take place. If 

 further evidence is required it may be given in the fact 

 that the up and down traces show, with hardly an excep- 

 tion, practically identical temperatures, especially in the 

 isothermal region, where, if a lag came into play, they 

 ought to be most divergent. 



With regard to ascending and descending currents, we 

 know from our experience with kites that such exist. 

 The pressure of the air must be continuous, but discontinuity 

 in the density may arise from an inversion, the tempera- 

 ture suddenly rising with the height. Five degrees centi- 

 grade is a not uncommon amount, but this means a 

 change of density of less than 2 per cent. The free lift 

 of a balloon in these ascents is about equal to the whole 

 weight lifted, and hence a sudden change of 5° C. may 

 mean a change of 4 per cent, in the free lift. To meet 

 this the square of the velocity must change by 4 per cent, 

 and the velocity by 2 per cent. This is of quite a different 

 order to the observed variations in the vertical velocity, 

 which in the lower strata may reach 50 or more per cent. 

 It would be of interest to know the rate at which the gas 

 inside approximates to the outside temperature, and the 

 amount of difference which may arise from solar insola- 

 tion. 



No doubt the point raised by Mr. Craig is an important 

 one, and may lead to serious errors in the height unless 

 especial care is taken to guard against it, but it cannot 



XO. 2045, VOL. 79] 



account for the observed phenomena. In England, ai 

 least, we have never claimed great accuracy in the re- 

 corded heights, and I do not believe such accuracy to bt 

 possible, but our traces show the commencement of thi- 

 isothermal part at 10 to 12 kilometres, and the balloons 

 often reach 18 kilometres. That elastic fatigue can 

 account for the space between 11 and 18 kilometres is 

 out of the question. ^^'. H. Dines. 



Curious Effect of Surface Ablation of a Glacier. 



DtJRiNG a traverse of the Corner Glacier last summer, a 

 peculiar feature in connection with the surface ablation 

 came to my notice which I have never seen described or 

 even referred to. I therefore venture to give a short 

 description of this interesting phenomenon. The surface 

 of the glacier is studded with the usual pools of clear 

 water, due to clusters of small stones which become heated 

 and sink some distance in the ice, the north and south 

 retaining walls of the pool sloping in a northerly direction. 



On the southern or sunny side of most of the pools 

 there was a spine-like projection of ice standing vertically 

 from the general surface, and tapering upwards to a jagged 

 knife-edge with an east and west alignment. Upon the 

 iiorthern rather flat (or slightly concave?) sides of these 

 spines there were sometimes vertical ridges, and at 

 irregular intervals the remains of thin, horizontal shelves 

 ol ice, which represented former levels of the water in the 



Spine-like projections 

 of small stones) i 

 18 inches. 



: upon the sunny side of pools (formed by clusters 

 ; Gorner Glacier, August, igo8. Height about 



pools. The sketch roughly shows this disposition. At tfie 

 lime there was no opportunity of making accurate measure- 

 ments ; many of the spines, however, were at least 

 18 inches high, and there appeared to be a proportional 

 relation between the height, width, and breadth of the 

 spine and the width and depth of the pool. Apparently 

 each projection is the relic of what was the southern wall 

 of the pool when the general surface of the glacier was 

 at a higher level (as shown by the horizontal shelves of 

 ice), and which by some means, perhaps partly because 

 that wall is in shade, has been maintained at a low 

 temperature, and thus enabled to resist the general surface 

 ablation. Bernard Smith. 



Blackheath. 



Moral Superiority ? 



Is this worthy of record as e.xceptional ? I fed the birds 

 with soft lumps of bread (not crumbs) this morning ; there 

 was a strong frost and 6 inches of snow. The first instant 

 arrivals were two redbreasts and a blue tit. All three 

 flew to the same lump of bread ; but the little tit turned 

 with fury on his two (ordinarily) combatant opponents, 

 drove them away, and — while I watched — took his fill of 

 the bread he had won until the usual crowd of sparrows 

 appeared, when he decamped with what was left of his 

 meal, and then the redbreasts — as always — ruled the roost. 

 F. C. Constable. 



Wick Court, near Bristol, December 30, 190S. 



