i88 



NATURE 



[April 14, 1910 



LETTERS TO THE EDITOR. 



[The Editor does not hold himself responsible for opinions 

 expressed by his correspondents. Neither can he undertake 

 to return, or to correspond -with the writers of, rejected 

 manuscripts intended for this or any other part of Nature. 

 No notice is taken of anonymous communications.] 



Colour of Water and Ice. 



I HAVE read with much interest Sir Ray Lankester's 

 letter on the colour of water in Nature of March 17. I 

 remember discussing this same problem in the case of 

 ice with Sir Joseph Thomson when we stood at the foot 

 of the great glacier at Glacier, B.C., during the western 

 excursion of the British Association last summer. The 

 rich blue colour of the hard, clear ice was remarkable, 

 even in quite small pieces. The same blue colour is noticed 

 when surface-ice, which has been formed slowly by con- 

 duction, is taken out of the St. Lawrence River. The 

 blocks lose their colour when they are exposed for long 

 to the light, and especially rapidly when exposed to sun- 

 light. Coloured sediment and air cavities in the ice detract 

 from the colour. I am inclined to believe that the colour 

 of ice is a real absorption effect, due to the large molecular 

 aggregates forming the structure, which absorb . the long 

 rays, and not a " blue sky " effect, as I suggested after 

 seeing the blue ice of the glacier. 



In the case of water, all the physical properties indicate 

 the_ presence of complex molecular aggregates in solution, 

 which_ become gradually reduced in number as the tempera- 

 ture rises. Thus the variation of specific heat, of density, 

 of viscosity, and compressibility, all disclose an effect due 

 to a gradual diminution of the molecular aggregates. 1 

 believe these are the same as the .ice molecules, and 

 constitute the absorbing medium which gives water its 

 blue colour. Sea-water is particularly blue, and here we 

 probably have added the effect of the salt molecules, in 

 addition to the fact that the water is very clear. 



Mr. W. H. Sherzer has shown (Smithsonian Report, 

 1907) that the blue colour of the water and ice of the 

 glaciers of the Canadian Rockies is a real absorption 

 effect. The blue colour is increased by the presence of 

 minute white sediment, but not by coloured sediment. 



If it has not been already tried, it would be very interest- 

 ing to see what effect temperature has on the greenish-blue 

 light transmitted through very pure water. If the c»lour 

 is due to the presence of ice molecules, it should grow 

 less as the numbers are reduced. I cannot help thinking 

 of the beautiful blue colour of liquid air as soon as most 

 of the nitrogen has boiled away ; if this were due to the 

 presence of complex oxygen molecules, such as ozone, it 

 would be somewhat similar to water. 



H. T. Barnes. 



McGill University, Montreal, March 29. 



convenience a rainfall of 36 inches, we might have the 

 three following distributions : — 



Centre of Gravity of Annual Rainfall. 



The question whether Mr. Cook's suggestions in Nature 

 of March 31 have a practical value can be very simply 

 settled. It is proposed to consider the month to month 

 rainfalls at a place as a series of parallel forces, 

 Pit p2y • ' Pi2J say, where the distance (X) from the 

 beginning of the year of the corresponding "centre of 

 gravity " is given by 



X- /i + ^A + 



+ 12/,, 



P1+A+ • • • +/]2 



Now, if we assign arbitrary values to any ten, say, of 

 the p's (and these ten p's could be selected in sixty-six 

 ways), then, the position of the CO. remaining the same, 

 we have obviously a single linear equation in two variables 

 to give us the values of the two remaining ^'s, and this 

 equation can be solved in an infinite number of ways. 

 Thus the same C.G. can be given by an infinitely varied 

 arrangement of sizes of the p's, and therefore its position 

 gives no indication whatever of the monthly distribution 

 of the rainfall of the places referred to. 



To illustrate by three simple examples. The absolute 

 value of the elements is of no importance, and taking for 

 NO. 21 1 1, VOL. 8s] 



January 



February. 



March 



April 



May 



June 



July 



August 



September 



October 



November 



December 



Total Rainfall ... 



C.G 



Rainfall Moment 



3 

 3 

 3 

 3 

 3 

 3 

 3 

 3 

 3 

 3 

 3 

 3 

 36 

 6-5 

 234 



36 

 6-5 

 234 



36 



6-5 

 234 



Thus A, B, and C have the same annual rainfalls, the 

 same C.G., and the same "rainfall moment." The ques- 

 tion whether or not the seasonal distributions correspond 

 to those of actual places on the earth's surface is not to 

 the point, though, as a matter of fact, C approximates to 

 the typical Levantine curve. 



Thus places with very different rainfall distributions may 

 have the same C.G. and the same rainfall moment, and 

 the proposed method of comparing the rainfalls of various 

 places appears to have neither a theoretical nor a practical 

 value. The method may possibly have a certain critical 

 value in comparing the yearly variations at a particular 

 place, where there is but little change in type from year 

 to year, and especially in such a country as India, where 

 the seasonal rainfalls are exceptionally well marked ; but 

 this seems doubtful, and in any case the method could be 

 used only in conjunction with the actual monthly values. 



Andrew Watt. 



Scottish Meteorological Society, Edinburgh, April 6. 



Certain Reactions of Albino Half. 



Under this heading in Nature of March 24 (p. 96) Miss 

 Igerna SoUas referred to some experiments of mine upon 

 the hair of albino rats, in which she failed to obtain one 

 of the reactions described in my note (Proc. Physiological 

 Soc, March 27, 1909). It is, of course, not improbable 

 that different albino rats may carry different chromogens, 

 and that some of them may lack the one which, when 

 oxidised with H^O^, gives a brownish colour. On the 

 other hand, the failure may be due to the presence of 

 some of the formalin, which may not have been completely 

 washed away from the previous reaction. 



There is one new observation which I should like to 

 record here, since it bears upon the HjOj reaction. The 

 action of H^O^ is an oxidising one, and the production of 

 a brownish tint may be interpreted as due to the oxida- 

 tion of a colourless chromogenous body present in albino 

 hairs. If this interpretation is right, other oxidising 

 agents should produce a similar tint. During November 

 of last year I casually placed two dead albino rats upon 

 the top of one of my cages, these latter being kept out of 

 doors. The rats were forgotten, and left exposed to the 

 air for about a fortnight. During this interval the weather 

 had been wet and warm for the time of year. Upon 

 discovering them at the end of this interval, I noticed that 

 on both rats the upper side, and part of the belly and 

 back, had assumed the same sort of brownish" tinge that 

 HjOo produces. The under side of both rats, which had 

 been protected from the wet and light by its contact with 

 the cage, was quite white. It thus seems possible to 

 oxidise the chromogenous substance ostensibly present in 

 albino rats by the oxygen of the air in the presence of 

 continuous moisture. I do not think that light played 

 much part in the reaction, since throughout this period it 

 was very dull weather. 



The mention of light brings me to another point in Miss 

 Sollas's note. She says that prolonged immersion in the 



