June 20, 1901] 



NA TURE 



LETTERS TO THE EDITOR. 

 [The Editor does not hold himself responsible for opinions ex- 

 pressed by his correspondents. Neither can he undertake 

 to return, or to co> respond with the writers of, rejeclei 

 manuscripts intended for this or any other part of Nati'ke. 

 Ao notice is taken of anonvmotis cominunicatiotis.'\ 



Does Chemical Transformation Influence Weight ? 



Careful experiments by Heydweiller, published in the last 

 number of Drude's Annalen (vol. v. p. 394), lead their author 

 to the conclusion that in certain cases chemical action is accom- 

 panied by a minute, but real, alteration of weight. The chemical 

 actions here involved must be regarded as very mild ones, e.g. 

 the mere dissolution of cupric sulphate in water, or the substi- 

 tution of iron for copper in that salt. 



The evidence for the reality of these changes, which amount 

 to 02 or 0"3 nig., and are accordingly well within the powers 

 of a good balance to demonstrate, will need careful scrutiny ; 

 but it may not be premature to consider what is involved in 

 the acceptance of it. The first question which arises is — does 

 the mass change as well as the weight ? The affirmative answer, 

 although perhaps not absolutely inconsistent with any well 

 ascertained fact, will certainly be admitted with reluctance. The 

 alternative — that mass and weight are not always in proportion 

 — involves the conclusion, in contradiction to Newton, that the 

 length of the seconds' pendulum at a given place depends upon 

 the material of w-hich the bob is composed. Newton's experi- 

 ment was repeated by Bessel, who tried a number of metals, in- 

 cluding gold, silver, lead, iron, zinc, as well as marble and quartz, 

 and whose conclusion was that the length of the seconds' pen- 

 dulum formed of these materials did not vary by one part in 

 60,000. At the present day it might be possible to improve 

 even upon Bessel, or at any rate to include more diverse sub- 

 stances in the comparisons ; but in any case the accuracy obtain- 

 able would fall much short of that realised in weighings. 



As regards Heydweiller's experiments themselves, there is 

 one suggestion which I may make as to a possible source of 

 error. Is the chemical action sufficiently in abeyance at the 

 time of the first weighing? If there is copper sulphate in one 

 branch of an inverted U and water in the other, the equilibrium 

 can hardly be complete. The water all the time tends to distil 

 over into the salt, and any such distillation must be attended 

 by thermal effects which w-ould interfere with the accuracy of 

 the weighing. Rayleigh. 



June 1 1. 



The National Antarctic Expedition. 



I.N consequence of a cable received yesterday from London> 

 telling me that the instructions for the conduct of the Nationa 

 Antarctic Expedition that had been passed by the Joint Committee 

 of the Royal Society and the Royal Geographical Society have 

 been greatly altered, I feel it my duty to resign the post of 

 head of the civilian scientific staff of the expedition, which I 

 had provisionally accepted. The organisation of the expedition 

 now passed leaves the head of the civilian scientific staff 

 nominally respon.sible for most of the scientific work of the 

 expedition, but gives him no power to secure the performance 

 of the scientific part of the programme. 



The responsibility for my withdrawal at so late a date rests 

 with those who have delayed until now the settlement of the 

 programme and organisation of the expedition, which .should 

 both have been decided, as I understood they had been, before 

 the ship and most of the equipment had been ordered. 



I trust the protest of my withdrawal will secure to my suc- 

 cessor more lavourable conditions of work than the altered 

 instructions would have given me. 



University, Melbourne, May 4. J. W. Gregory. 



The Settlement of Solid Matter in Fresh and 

 Salt Water. 

 Since the publication of the report of Mr. Slidell' on the 

 deposits of tire Mississippi delta, containing the remarkable 

 statement that while the deposit contained in the river water 

 of the Mississippi took from ten to fourteen days to settle, with 

 solutions of salt, sea water or sulphuric acid the water became 

 limpid in from fourteen to eighteen hours, it has generally been 



1 Report on the Mississippi River by Humphreys and .Abbott, 1861. 

 KO. 1 65 I. VOL, 64] 



taken as an accepted fact that alluvial matter settles more 

 rapidly in salt than in fresh water. Sir Archibald Geikie, in 

 his "Text-book of Geology," endorses this theory; and in a 

 recent article in the American Engineeriw^ Magazine on the 

 transportation of solid matter by rivers, Mr. Starling, one of the 

 Government river engineers, states that a small quantity of salt 

 or other foreign material dissolved in water will diminish the 

 suspending power and increase the rapidity of subsidence to a 

 marked degree, sometimes even many hundred-fold. 



On the face of it the result naturally expected would be, that 

 as sea water is of greater specific gravity than fresh water, and 

 more viscous, the grains of solid matter would sink more slowly 

 in salt than in fresh water. The very great distance over which 

 solid matter brought down by rivers remains in suspension after 

 reaching the sea, extending from six miles from the mouth of the 

 Rhone to thirty-five from the outlet of the Nile, up to 300 miles 

 over which the sea water is stated to be discoloured by the 

 effluent of the Amazon, appears to indicate that salt water is 

 capable of retaining solid matter in suspension for a longer time 

 than fresh water. 



Experiments made by Mr. Vernon Harcourt with alluvial 

 matter placed in suspension in sea water and fresh water, and in 

 solutions containing different strengths of salt and other foreign 

 material, although not of a conclusive character, show that there 

 is little difference between the rate of deposit in sea or in fresh 

 water. Of samples from different estuaries which were allowed 

 to settle in sea water and pond water respectively, the particles 

 of the former took about 9 per cent, more time to subside than 

 the latter. The general conclusion he arrived at was that, 

 though sea water promotes the deposit of "very light clayey 

 matter contained in river silt under favourable conditions, there 

 are no grounds for regarding it as exercising the very preponder- 

 ating mfluence on the formation of deltas attributed to it by 

 geologists."' 



The writer some time ago investigated this subject in connec- 

 tion with researches he was then making as to alluvial deposits 

 in estuaries, and has again more recently conducted a series of 

 experiments the mean results of which are given in the following 

 table : — 



Table Showing Kate of Settlement of Solid Matter in Fresh 

 and Salt Water. 





Whiiing 

 Plaster of Pari> 

 Warp, Trent 



( Fine Warp, \ 

 ( Dutch Riv.r i 



Silt, Salt Marsh 



Warp. do. 

 r Alluvium, Bos-l 

 I ton Dock .../ 



do., R. Parrett 1 



do., Tilbury Bsn. 



Brick Clay ... 



It will be seen from this table that the rate of deposit depends 

 on the minuteness of the panicles in suspension, and varies 

 nearly in proportion as the square of the diameter of these. 



With sand and silt there was practically no difference in the 

 rate of settlement in fresh as compared with salt water. When 

 the particles of the solid matter were very fine, as in the case 

 of what is generally known as mud or ooze, the rate of settle- 

 ment was slightly more rapid in salt than in fresh water ; but 

 there was nothing to justify the conclusion arrived at by Mr. 

 Slidell. 



All the material was first screened through a sieve having 

 ninety meshes to the lineal inch. 



The proportion by weight of solid matter to water was that 

 which was found to exist on the average of fourteen large rivers 



1 '' Investigatior 

 of River Silt '■ (Mh 



on the Action of Sea Water in .decelerating the Deposit 

 . Proc. Inst. C.E., vol. cxlii.). 



