Nov. II, 1880J 



NATURE 



35 



])03sible. But what is still more puzzling is how I'rof. CarneUcy 

 succeeded in burning his fingers witli the ice. Our previous 

 knowledge would lead us to suppose tint tlie outside surface of 

 the block of ice was a. free surjacc, and tint therefore it would be 

 impossible, however high the temperature of the inside of the 

 block, to heat the outside above the " melting point," as we 

 should expect the ice to melt or to sublime at the outside, and 

 keep the temperature at 0° C. 



These expectations being disappointed, we naturally look to 

 the decreased pressure under which Prof Caniclley's experiments 

 were made for an explanation of this most unexpected state of 

 matters. Now it is very evident that when dealing w ith pressures 

 of about one atmosphere, and with temperatures of 120' and 

 180° C, that pressure, as pressure, has nothing wh.atever directly 

 to do with the " melting point " of the ice. While this is the 

 case, it is equally evident that it has a most important influence 

 on the surroundings of the ice. At the jircssure of 4*6 mm., at 

 which the experiments were m.ade, no water would be present, 

 there would l)e nothing but ice and water-vapour. Here then 

 appears to be the great leacliing of Prof. CarncUey's experiments. 

 They show that the siirfue of ice bounded by its own vapour is not 

 a "free surface." Tliis result is so very unexpected that much 

 consideration will be necessary before we can re arrange our 

 ideas to meet the new facts. 



We might imagine that nothing could be more free than the 

 surface of a body bounded by nothing but its own vapour, yet 

 I'rof. Carnelley's experiments seem to say it really is not so, and 

 not being a "free surface," we of course know nothing whatever 

 of how high the temperature will require to be before the ice 

 will melt under these conditions. 



These ex]>eriments of Prof. Carnelley's are so interesting that 

 we w.ait with impatience a full description of them. His results 

 indicate something new with regard to the influence of a liquid 

 on its melting solid. I observe that Prof. Carnelley's results are 

 doubted by most of your correspondents, but for the present we 

 must accept them when Prof. Carnelley distinctly states that 

 the temperature of the ice was taken by means of a thermometer 

 in contact with the ice. John Aiticen 



D.arroch, Falkirk, N.B., October 30 



Wire Torsion 



In the letter in Nati'RE, vol. xxii. p. 604, which we wrote at the 

 request of M.ajor Herschcl, who asked for information regarding the 

 connection between tensional and torsional strains of a brass wire, 

 we mentioned that there were many papers scattered througli the 

 Proceedings of learned societies dealing with the fluidity of metals. 

 There is one communication to which we might specially have 

 referred, .as it deals in particular with the torsional yielding of 

 wire; under tension, and this is a paper on " Torsion," by Prof. 

 G. Wiedemann, in the .■Iniialen der I'hysil; und Chemie, No. 4, 

 vol. vi., 1879, pp. 485-520, and of which a translation is given 

 in the Philosophical Magazine, vol. ix., January iSSo, pp. i-t5, 

 and I'ebniary, pp. 97-109. The first jiart of this paper gives a 

 detailed account of experiments which show ; — (i) That a brass 

 wire often subjected to a particular torsion, either in one or in 

 both directions, becomes "killed " for any less torsions, that is, 

 follows Ilooke's law for its temporary torsions; (2) that a wire 

 under tension acquires greater tor-ional set from a given torsional 

 couple than when the wire is i;nextended ; (3) that a wire under 

 even considerable tension may be killed by torsion in alternately 

 opposite directions, that is, it w ill obey I looke's law for any tension 

 or torsion less than the .stresses .actually applied originally. Prof. 

 Wiedemann in the second part of his i)aper considers the well- 

 known "agitation effects," and enters on an explanation of the 

 phenomenon based upon molecular allineations referring to the 

 magnetisation theory of Weber and Kolrausch which is based on 

 the same idea. 



The strains in Prof. Wiedemann's wires were however much 

 ess than in those used in Major Ilerschel's experiments. 



John Pekhy 



London, November S W'. E. Ayuto.V 



Heat of Formation of a Compound 

 In Nature, vol. xxii. p. fioS, there is a jiaper on "Recent 

 Chemical Research," in which under the head of work by 

 Thomsen the following law is enunciated : — 



"The heat of formation of a compound substance is the differ- 

 ence between the sum of theheats of combustion of the constituent 

 elements of the compound and the heat of combustion of the 



compound itself." After that it is shown that this is not the 

 true heat of formation of the compound, as many important 

 corrections have to be made. On referring to Pertliclol's "Essai 

 dc Mecanique Chiuiique " I find the following :—" The heat 

 of formation of .an org.nnic compound from its elements is the 

 difference between the sum of the heats of total combustion of 

 its elements and the heat of combustion of the compound with 

 formation of identical products." 



Can any of your readers inform me whether Thomsen or 

 Berthelot first enunciated this law ? 



Another point is, that Berthelot apparently makes no refer- 

 ence to the corrections for the he.at ai)sorbed in dissociating the 

 molecules of the elements, &c. A. P. Eaukie 



Edinburgh, November i 



The Yang-tse, the Yellow River, and the Peiho 



In replying to the letter of your correspondent (Nature, 

 vol. xxii. p. 559) on the subject of my recent (laper on these 

 three rivers, 1 have to thank him for his very probable explana- 

 tion of the excessive estimate made by Sir George Staunton of 

 the amount of sediment discharged by the Yellow River. 



The estimate given in my paper of the water-discharge of the 

 River Plate is ipso facto an assuin|)tion m.idc by Mr. George 

 Iliggin from Mr. Bateman's calculation of the minimum flow of 

 that river, which he found to be 670,000 cubic feet per second. 

 It might liave been belter, however, if 1 had added Mr. Higgin's 

 qualifying remark that such an estimate of the mean volume of 

 water was "very much under the mark" (Nature, vol. xix. 



p. 555)- 



'1 lie anomaly of the surface current varying in velocity witli 

 the same average depth of water has not been unnoticed by 

 myself, though I am unable to give a satisfactory explanation of 

 the difficulty. H. B. Gurry 



Woodlane, Falmouth, November 6 



The Thresher 



W'llAT is the "thresher"? It is generally assumed to be the 

 fox-shark {Atopias vulpes), but in a recent number of Land and 

 Water — which I have only just seen — Mr. Fiv.nl; BucUland s.ays 

 that he believes it to be "the gladiator dolphin or sword 

 grampus" (0;-(-« 'gladiator). This he infers from a dr.awing of 

 Lord A. Campbell's, of which he gives a copy. The tail, he 

 .says, is not that of the fox-shark. But as it is helerocercil it 

 cannot be that of a gramjius or any other Cetacean. Whatever 

 it is I suppose that there is no doubt that it throws itself out of 

 the water (" high as the masthead " [of a trawler] one of Mr. 

 Buckland's correspondents avers). Does it do so more than 

 once ? Once, many years ago, between Sydney and New Zealand, 

 I saw, what they said was a fight between a thresher and a 

 whale, but there was nothing to be seen beyond a splashing of 

 the water. Last year off Lisbon I witnessed a similar event. 

 Docs the sword-fish also attack the whale ? Lord A. Campbell, 

 in the letter accompanying his drawing, estimated the length of 

 his thresher at " upwards of thirty feet ; " this is twice the length 

 given by \arrell. Francis P. PAscoii 



Cctolier 30 



Since the above was written I see that Dr. Giinther, in his 

 new work on Fishes, says : " Statements that it (the fox-shark) 

 has been seen to attack whales and other large cetaceans rest 

 upon erroneous observations" (p. 322). 



"Student" should refer to Newcomb's "Popular Astro- 

 nomy " w ith respect to the larger telescopes. For results he must 

 refer to the publications of the Royal and Astronomical Societies, 

 the Washington Observatory, &c. 



Paul LAi"ARf;uE. — We regret we have no further details on 

 the labours of the U.S. Fish Commission in increasing the food 

 supply of the country. 



TLLU.STRA TION.S OFNE IV OR RA RE ANIMALS 

 IN THE /.OOLOCICAL SOCIETY'S LIVING 

 COLLECTION 



I. 



Till', sagacious founders of the Zoological Society of 

 London made it a special rule that no dividends or 

 gilts of any kind should be distributed amongst the 



