October 17, 1895] 



NA TURE 



597 



t. ,.iiilii>iiHi. liai -.ciiijua doubt having been raised on this poim 

 by so higli an authority as M. le Chatelier, we have thought it 

 right to make further experiments. 



These experiments convince us that the freezing point of 

 molten silver is lowered and rendered variable when the surface 

 is exposed to the air. We also find that by blowing oxygen 

 through the molten metal, the absorption of this gas is sufficiently 

 great to lower the freezing point 20°. Moreover, when the 

 oxygen is removed by the action of either carbon, coal gas, or 

 hydrogen, a constant maximum freezing point is reached. 

 Further, if the atmosphere of hydrogen, or coal gas, be replaced 

 by carbon dioxide, there is no change in the freezing point, 

 whilst if nitrogen be used to sweep out the hydrogen, there is a 

 slight fall. In neither case does the removal of the hydrogen 

 bring about a rise, as should be the case on M. le Chateliers 

 hypothesis. 



Another strong reason for believing that the true freezing point 

 of silver can only he obtained in a reducing atmosphere, is to be 

 found in the remarkable constancy with which a considerable 

 mass of pure silver maintains its temperature from the moment 

 that freez'ng commences until the whole is solid, provided it has 

 not l)een exposed to the action of free oxygen. It is also note- 

 worthy that in a reducing atmosphere the melting and freezing 

 points arc identical. 



In)pure substances do not as a rule behave in this way, 

 and hence it is improbable that the silver can contain dissolved 

 hydrogen. In an oxidising atmosphere the freezing point is 

 less sharply marked, and the silver behaves as if it were impure. 



These are our leasons for venturing to differ from M. le 

 Chatelier, and we hope that he will further examine the question. 



Cambridge, October 12. C. T. Hevcock. 



F. H. Nevm.i.e. 



The Genus " Testacella.' 



Plant-Animal Symbiosis. 



In your issue of August 22, 1895, ^'f' Schwarz describes his 

 finding in South Africa some ants inhabiting the thorns of a 

 mimosa tree, by which he evidently means a species o{ .Uacia. 

 This symbiosis is well known out here, and probably also in 

 Furope, as will be seen by a reference to Schimper's " Wechsel- 

 bezichungen zwischen Pflanzen iind .Xmeisen im tro]iischen 

 Amerika," p. 48. I first observed ants inhabiting the thorns of 

 Acacia horrida in the neighbourhood of Grahamstown about six 

 years ago. I also found them near I'ort Alfred. As far as my 

 repeated observations go, the partnership between the ants and 

 the trees is a very one-sided one. 



The former receive shelter and food from the trees, whereas I 

 have failed to find that the latter derive any advantage from it. 

 This last conclusion is not surprising, as, firstly, amongst the 

 " mimosa "-scrub near Grahamstown, one only finds here and 

 there a tree the thorns of which are inhabited by ants, and as, 

 secondly, in some years all individuals of Acacia honida are 

 completely denuded of their foliage over wide areas by cater- 

 pillars, ^loreover the ants (of which I found two different 

 kinds) are, as Mr. Schwarz rightly o'.)serves, not at all aggres.sive, 

 whereas Belt showed that the little ants living in the hollow 

 thorns of Acacia spJucrocephala in Central .\merica are very 

 pugnacious, and protect the plant against browsing mammalia 

 and insect enemies. 



The two ca.se3 are, therefore, very different from one another. 



S. ScHoxi.AXn. 



.\lbany Museum, Grahamstown, South .\frica, 

 September 16. 



The Recent Dry Weather. 



Willi reference to the recent remarkable weather, both at the 

 commencement of the year and during Sejitember, it is worth 

 while calling attention to the climatological period of about 

 thirty-five years, which I'rof. Briickner, of Berne, pointed out 

 as existing relatively to (he years or groups of years characterised 

 by marked cold or heat, as mentioned in vol. xliii. p. i6j of 

 N.ATfRK. He therein indicated the years 1700, 1740, 17S0, 

 1815, 1S50, and 1880 as centres of cold periods, while the years 

 1720, 1760, 1795, iSjo, i860 (and now 1895) appear ^s centres 

 of warm, dry periods. The coincidence for the present year is 

 certainly remarkable, and merits attention as to the causes which 

 underlie these periodic fluctuations of weather. 



Dublin, October 11. J. P. O'Rlcii.l.v. 



N-0. 1355, VOL. 52] 



In Nature for last year the writer gave a list of the localities 

 for Testcuella sciilii/iiiii which had come under his notice. With 

 a view to making this list more complete, and to obtaining a 

 more definite idea of the distribution of the various species of 

 the genus in the British Isles, the writer would be greatly in- 

 debted to any reader of X.vtlre who could forward to him. 

 localised specimens of Testacella, alive, or preserved in alcohol, 

 the present month being a likely one for the coming above 

 ground of these slugs, which should now be found under logs 

 and stones in the neighbourhood of rich garden soil. 



WiiKKK:> Mark Wki:k. 



" Holmesdale," Brentwood, Essex. 



The B.A. Committee on Coast Erosion. 

 I N the reference, in your number of (Jet. 3, to ' • ( Icolog)- at the 

 British Association," the statement as t.o the Coast Erosion 

 Committee, in their final report, recomiHending a " De|>art- 

 mental Committee of the House of Commons," ro inquire into 

 the subject, is taken from the " first proof" of the report, which 

 was drafted by myself as surviving Secretary. The suggestion has 

 not been adopted by the majority of the Committee, who con- 

 sidered their duty did not extend to drawing up and formulating 

 recommendations. This termination I regret, as when the 

 ^Vssociation adopted my suggestion in 1881, to appoint this 

 Committee, I hoped it would have had a practical outcome, 

 leading to the conservation of our coasts. 



Chari-Es E. De R..VXCE. 



A Substitute for Sulphuretted Hydrogen. 

 In your Notes of February 14 last, you state that ammonium 

 thio-acetate has been found to be a satisfactory substitute for 

 sidphuretted hydrogen in chemical analysis. Can any of your 

 readers tell me where I can obtain it ? I cannot find it in 

 catalogues of chemical manufacturers. RusTiciJS; 



THE GRAPHICS OF PIANO TOUCH. 

 TX/I UGH trouble has been taken in order to construct 

 ■'•*■*■ an apparatus that will leprodtice graphically the 

 effects of touch in keyed musical instruments. The 

 experiments are most easily made with the piano, and 

 have therefore been tried on that instrument. 



Recently a most interesting article appeared in 

 the Revue Scie/ilijique, written by M.M. liinet and 

 Courtier, who ha\ e studied this subject closely, and have 

 made many experiments with their apparatus. They have 

 treated the matter very fully in their article, of which the 

 following is a resume' : — 



When a certain point of perfection has been attained 

 in piano playing, it becomes \ery hard to distinguish 

 inequality of touch : yet, owing to the \arying strength 

 of the fingers, it is only with much practice that perfect 

 equality is possible. .As will be seen further on, involun- 

 tary mo\cinents and irregularities, scarcely ]jerceptible to 

 the ear, are shown b\- the graphical method. 



The apparatus { Fig. 1) is quite simple in construction, 

 and consists chiefly of an india-rubber tube, placed under 

 the key-board, united at its two extremities by a register- 

 ing drum, also of india-rubber. When the notes of the 

 piano are played, the pressure on the tube causes a wave 

 of air to be sent through it into the drum, upon which is 

 attached a pen that in the ordinary way is made to 

 record its movement on a moving roll of paper. The 

 wa\c makes the drum vibrate, which in its turn jerks 

 the pen, thus causing irregular marks to be left on the 

 paper. The board on which the tube rests is regulated 

 b\- means of « edges adjusted by a screw, the board being 

 either lowered or raised. When raised it almost reaches 

 the notes of the piano, and in this case the registering 

 action takes place : but if it is lowered, the whole appar- 

 atus is disconnected from the key-board. 



When no notes are being played, and the registering 

 drum is connected, i.e. the board is raised, merely a 

 straight line is drawn. In Fig. 2, first a is struck, then 

 two notes with b, then three notes with c, and so on. It 



