NA TURE 



lNovembek 14, 1901 



depending on the height of the barometer. M. Vallot 

 adopts the same form of expression, but unfortunately 

 does not give the means of conparing his numbers with 

 those found in the earlier investigation. Moreo,-er, the 

 methods employed are not quite the same, and a slight 

 difficulty arises fro 11 the selection of a silver salt whose 

 maximum effect is not exhibited at the same part of the 

 spectrum as in the case of Bunsen and Roscoe's inquiry. 

 Making allowance for a slight discrepancy in this respect, 

 we gather that while Bunsen and Roscoe claimed that 

 59 per cent, of the solar light was lost by transmission 

 through the atmosphere, M. \'allot finds that only 40 per 

 cent, is so lost. Treating the sensitised paper in such a 

 way as to make the maximum sensibility coincide with 

 the optical maximum, M. \'allot finds the coefficient 

 of atmospheric transmission to be 0826, corresponding 

 with the mean of the values found by Bouguer, Leiden 

 and Trcpied. While the agreement quoted is quite 

 satisfactory, we miss any reference to the work of 

 Captain Abney on the Faulhorn 1 2683 m.), of Dr. Muller 

 at the Etna Observatory (2944 m.), or of Prof. Langley 

 on Mount Whitney (3513 m.). These latter authorities 

 have all found a rather larger transmission coefficient 

 than others who have worked nearer the sea le\el, and 

 the greater altitude of the Mont Blanc Observatory 

 might have afforded some explanation. Doubtless if the 

 object be to inquire into the effect of the variation of a 

 particular factor, there is an apparent advantage in con- 

 ducting the observations in such a way that that factor 

 is most materially affected, but the advantages can be 

 considerably discounted by the extra difficulties intro- 

 duced by the remoteness of the situation and other 

 causes, and though M. Vallot and his assistants have 

 striven manfully with the difficulties, greater weight 

 would attach to his results if the observations had been 

 more frequently repeated, though at a lesser height and 

 with a smaller diminution of the barometric pressure. 



The means employed for discussing the relative 

 velocity of water in streams and torrents is discolouration 

 of the water by means of a powder (fiuorescin 1, which on 

 dissolving rapidly gives rise to a green tint, and then 

 observing the tmie at which the discoloured water arrives 

 at a distant station. M. \'allot very properly insists 

 upon the necessity of a sufficient expenditure of the 

 powder, which can be determined experimentally accord- 

 ing to conditions varying with the quantity of outflow that 

 passes a given point, the colour of the water, &c. 

 Observations made in this way do not permit of very 

 great accuracy ov/ing to the tendency of the coloured 

 water to spread itself according to the character of the 

 channel over which it passes, but when it is a question 

 of determining the velocity in a subterranean channel, 

 under glaciers, it is difficult to see what better means 

 could be devised. M. \'allot has confined his attention 

 mainly to the effect of the slope of the river bed on the 

 rate of flow, and concludes that this is most rapid when 

 the incline is about three in a hundred. As the slope 

 increased, contrary to his expectation, the rate as 

 measured horizontally, diminished. From this he con- 

 cludes that the onward rush of water along steep ravines 

 is more apparent than real. When the incline is very 

 steep the effect is to hollow out the bottom, giving rise 

 to great irregularities in depth which produce eddies 

 and tend to stop the hurrying progress. It is there- 

 fore the mean velocity with which M. Vallot is con- 

 cerned, and the distances measured vary from iodd to 

 4000 m. The maximum velocity observed in the uncon- 

 fined stream is 225 m. per second, and for a subterranean 

 current of the same slope the progress is atjoal one half 

 of this quantity. The same ratio obtains between open 

 streams and sub-glacier currents, whether the horizontal 

 or vertical velocity be measured. 



But the particular investigation to which M. Vallot 

 has devoted the greatest attention is the movement of the 



NO. 1672, VOL. 65] 



glacier, Mer de Glace, in its various parts. Here the author 

 enters into a very difficult inquiry, and it would be sur- 

 prising if some of his results did not differ from generally 

 received opinions. .Xs a contribution to a more complete 

 investigation the study made by M. X'allot is no doubt 

 valuable, and much of his work, such as the difficult 

 triangulation of the Mer de Glace, will be appreciated by 

 later obser\ers. But a complete theory of glacier move- 

 ment is not to be derived by a study. howe\ er minute, 

 of a single glacier in the space of a few years. .An 

 international commission under AL Forel has been at 

 work for a considerable period and is |)robably still 

 collecting data whicli show that the problem possesses 

 many perplexing variations, not the least dilficult being 

 the evidence of periodicity in glacier movement con- 

 nected with some obscure law that appears to affect the 

 general climate of the earth. During the eight years 

 that M. \'allot has been at work on this subject, he 

 claims to have established the following prominent facts, 

 which undoubtedly ha\e reference to the particular phase 

 of the motion which obtained in the period to which his 

 observations have reference. We regret that it is im- 

 possible to enter with minuteness into the character of 

 the evidence by which his conclusions are supported, but 

 they may be briefly summarised thus : 1 i) That the 

 progress of the glacier throughout the year is constant, 

 the summer does not hasten nor does the winter witness 

 any abatement of tlie uniform progress. Changes of 

 incline of the bed on which the glacier moves will always 

 explain any observed variation of velocity. ''2'i That the 

 uniformity of the velocity in all seasons is opposed to 

 any theory of regelation, or, indeed, to any explanation 

 in which changes of temperature play a part. (31 That 

 I the movement of the glacier does not partake of the 

 character of a \ iscous fluid, the whole moving as one 

 piece. i\L \'allot thus sums up the result of his long and 

 arduous labours, which have been pursued under great 

 difficulties with considerable skill and over a large area. 

 We can only hope that the experience he has gained v/ill 

 be still longer employed in this species of investigation. 

 " La conclusion de ce travail est que la progression dcs 

 glaciirs est caiisee par !c glisseineiit de la masse, sous 

 faction tie la pente dii lit, aidee par la poiissJe des parties 

 posterieures. La pesanteur seule parait etre en jeu, k 

 I'exclusion de toute action calorifique.''' 



GEOLOGY AND METEOROLOGY. 



T^HE subject of climatic changes has always been of 

 -'■ absorbing interest to geologists, and they have been 

 perhaps more puzzled to account for the occurrence of 

 plants of temperate or even subtropical character in 

 Arctic regions than for the occurrence of wide-spread 

 Arctic conditions in temperate regions To explain these 

 changes in the northern hemisphere, the alterations in the 

 distribution of land and water and consequent influence 

 on the Gulf Stream, the modification of the internal heat 

 of the earth, changes in the position of the earth's axis, 

 variations in the amount of heat given off by the sun, the 

 eccentricity of the earth s orbit, and even fluctuations in 

 the amount of carbon dioxide in the atmosphere, have 

 individually or collectively been invoked. Tne influence 

 of ocean currents, as modified either by the elevation of a 

 tract of islands to form a continental area or by the total 

 or partial submergence of a continent, has naturally 

 been regarded as of very great importance. Moreover, 

 the effect which such changes would have on winds has 

 not been neglected, although their local influence has not 

 been fully realized. 



The apparently wide extent of tropical and sub- 

 tropical climates during past epochs, with evidence of 

 progressive diminution in temperature in later Tertiary 

 times, has been held to be due to astronomical rather 



