i6 



NA TURE 



[February i, 1894 



The method followed was to find the distinctly preglacial 

 valleys tributary in the preglacial stream valley, now deepened 

 by glacial erosion, and occupied by Lake Cayuga. There are 

 several of these with the broad valley and gently sloping sides 

 so characteristic of old valleys, and so different from the steep 

 side post-glacial gorges. Their directions are at all angles with 

 that of the ice motion. If the main valley (the present lake 

 viUey) has not been sensibly deepened by the ice, these tribu- 

 taries should not be rock bottomed. In the present instance 

 the entire valley is found to be underlain by the Devonian shale 

 in place (nearly horizontal) ; and the bottom of the preglacial 

 valleys are there> at the lake margin, found to be over 400 feet 

 above the deepest point of the lake. 



There are three possible explanations of this phenomenon : 

 either (l) the rivers had a fall of over 400 feet in less than a 

 mile, while above this the slope was only very slight, or (2) 

 the lake valley has subsided 400 feet, or (3) it has been deepened 

 by ice eroion. Few will, I think, consider the first two to be 

 possible, and there is evidence that they are not. For fuller 

 details reference may be made to my forthcoming paper. 



It seems to me that we have here a reasonable and possible 

 method of testing the value of the rock basin theory, and I 

 believe that its application in other regions will show that ice 

 can, where conditions were favourable, excavate lake basins of 

 large size, and has done so. This conviction comes to me in 

 spite of distinct preconception and prejudice against the theory. 



R. S. Tarr. 



Cornell University, Ithaca, New York, January 15. 



Glacial Erosion in Alaska. 



References in your recent correspondence to my estimate 

 of the rate of erosion by the Muir Glacier in Alaska, call for a 

 supplementary statement. The estimate was made in 1886 by 

 determining the amount of sediment per gallon brought down 

 by one of the sub-glacial streams, and calculating as best I 

 could the area of the glacial basin, the amount of annual rain- 

 fall, and the probable waste by evaporation and by the formation 

 of icebergs. The result obtained was the removal of one-third 

 of an inch of rock per annum over the glaciated area. 



Since my visit to the Muir Glacier, Prof. H. F. Reid has 

 spent two summers on the same ground with more ample pre- 

 parations for collecting the facts. His report may be found in 

 the National Geop-aphic Magazine (Washington), vol. iv. p. 51. 

 According to his calculation, the erosion amounts to three- 

 quarters of an inch, or nearly three times as much as I had 

 estimated. I have little question that Prof Reid's estimate is 

 more nearly correct than mine, since my calculation was based 

 upon the removal of sediment from the entire drainage area of 

 the glacial amphitheatre. Prof Reid, however, rightly con- 

 cludes that this is full twice as large as the actual t)ed of the 

 glacier to which the glacial erosion was practically limited. 

 Making that correction, our estimates are in close agreement. 



It should be observed, however, that these observations do 

 not bear directly upon the question of the erosion of lake basins 

 by glaciers ; for the Muir Glacier, whose sediment was 

 estimated as above, is moving down a slope of about 100 feet 

 per mile. The erosion over this slope, therefore, may be quite 

 different from that at the foot of the glacier as it descends below 

 the water-level into the head of the tidal inlet, where, I should 

 presume, the erosive power would be soon reduced to a small 

 quantity. Still, the mechanical problem involved in calculating 

 the distribution of the force of a descending glacier as it reaches 

 the foot of the incline is too complicated for ready solution. 

 That there is some scooping out of a rocky basin in such cases 

 seems amply proved by the facts which I have quoted in my 

 " Ice Age "(pp. 237-239), from Prof. I. C. Russell, concerning 

 the formation of cirques in the Sierra Nevada Mountains. 



In my own observations two or three years ago, however, 

 upon Lake Geneva in Switzerland, I was led to believe that 

 whatever might be true of glacial erosion, attention enough had 

 not been given to the theory of a possible buried outlet leading 

 past Mount Sion and Frangy to Seyssel. Certainly the course of 

 the Rhone across the spur of the Jura Mountains at Fort De 

 I'Ecluse is very suggestive of recent occupancy. Among the 

 great lakes of America there can be but little doubt that Lake 

 Erie and some others owe their existence almost wholly to the 

 choking up of preglacial outlets of the valleys by glacial debris. 

 The great depth of Lake Geneva, however, would render it 

 improbable that it was wholly due to such a cause, and I dp not 



NO. 1266, VOL, 49] 



know that the conditions are such as to permit the supposition 

 made above. I distinctly remember, however, that from the 

 vicinity of Seyssel there was an unobstructed view between the 

 mountains towards Geneva, and that the gravel deposits extend 

 from Geneva far down towards those which appear a^iout the 

 head-waters of the small tributary to the Rhone which joins it 

 at Seyssel. Perhaps this theory has been fully considered and 

 refuted ; if so, I have not seen the refutation. 



G. Frederick Wright. 

 Oberlin, Ohio, January 17. 



On the Equilibrium of Vapour Pressure inside Foam. 



It is known that the vapour pressure near a curved liquid 

 surface is different from that near a flat surface, being less near 

 a concave surface and greater near a convex surface than near 

 a flat one. Now, inside foam bubbles the surfaces are approxi- 

 mately flat, except where three bubbles join to form an edge, and 

 along these edges the surface inside any bubble is concave with a 

 very small curvature. How does it happen that equilibrium caft 

 exist with the small pressure in these corners, and a larger pressure 

 required near the flat surfaces? In the first place it maybe 

 that equilibrium cannot exist, and that all foam is essentially 

 unstable ; and it would be almost impossible to disprove this 

 by a direct experiment. If, however, foam can be stable, it 

 seems as if the only conclusion possible were that the flat 

 surfaces will evaporate and thin down, the liquid condensing in 

 the corners, until the flat parts are so thin that they are in 

 equilibrium with a smaller vapour pressure than a thick liquid 

 surface would require. In other words, that the vapour pressure 

 near a very thin film may be less than it can be near a thick 

 one at the same temperature. It is evident that inside or out- 

 side a solid box stability must be possible, so that the second 

 alternative is the only solution. 



We see a phenomenon of this latter kind in the hygroscopic 

 films that cover glass. Being due to an attraction of the glass 

 for water there results what I am describing, namely, that in 

 an atmosphere incompletely saturated a film of such a thickness 

 can exist that the vapour pressure near it is such as corresponds 

 to the existing vapour pressure in the surrounding atmosphere. 



If we knew the connection between the thickness of a film 

 and the vapour pressure near it, it would be possible 

 to calculate the shape of a bubble near a corner. The pressure 

 at any point being that due to the thickness diminished by an 

 amount proportional to the curvature. So that \{ f {y) be the 

 pressure due to a film of thickness y and r be the radius of 

 curvature of the surface of the liquid near a corner, we get as 

 the equation of the surface 



T 



/O') - ,s, = constant, 



r(5 - a) 



when T is the superficial tension which may be a function o\ 

 y and 5 the density of the liquid and a of the vapour of the 

 liquid, which latter will depend on the vapour pressure inside 

 the bubble. To determine this and the constant we should 



[ know how much liquid we have at our disposal (V = \ydx) to 



lie round the bubble and to supply vapour inside, and it would 

 appear that inside a cubical box falling freely, for instance, 

 I a bubble would always be spherical unless the quantity of 

 liquid were sufficiently small to require the sides of the bubble 

 to be flattened against the sides of the box ; i.e. unless the 

 volume of liquid were less than i - 7r/6 times the volume of the 

 box. 



In connection with my letter in last week's Nature, I may 

 suggest that a possible cause of warming of solid powders when 

 mixed with liquids, is that the solids have already got a film of 

 water over their surfaces, which being on the outside in contact 

 with air, is at least on its outer surface in tension, and that this 

 enormous area of air-liquid film disappears when the solid is 

 immersed in liquid, and that the heat is due to the extinction of 

 this great film. It would require very careful measurements of 

 the heat evolved and estimates of the area of the film con- 

 cerned, to decide whether this would account for the heating: 

 observed. In my former letter I assumed that liquids would 

 soak up into dry powders, and that these latter warmed liquids 

 when mixed with them. The suggestion I am now making, 

 would account for a warming being produced by damp powders^ 

 or by spray or cloud falling into a liquid. 



G. F. Fitzgerald. 

 Trinity College, Dublin, January 29. 



