igo 



NATURE 



[Junk 20, 1895 



ce, boiling water and boiling sulphur. In most cases the ther- 

 mometer constants were determnie<l after the wire had been 

 ex[X)se<l to the action of the hot gases for about half an hour. 

 One wire, however was calibrated before Ix'ing used, and an un- 

 usually high value was obtained for the coefficient 5. After this 

 w ire had been ex|x>.sed to the hot gases, the value of S fell, how- 

 ever, to the nomial. The author thinks the abnormal value 

 may have been due to the fonnation of a gold platinum alloy 

 during the process of attaching the h ire to the leads, and that 

 this alloy was subsequently swept off by the hot gases. The 

 ]>a|X'r includes a number of tables and curves which embody the 

 numerical results, and show that concordant results can be ob- 

 tained on different days and with different thermometers. Prof. 

 I'erry said that an instrument for quickly recording varying 

 temperatures was greatly required by engineers. The tempera- 

 ture just inside the cylinder walls was, however, the most im- 

 portant to determine, and a knowledge of how the lemi>eralure 

 from I to 2 m.m. inside the walls varied would be of the greatest 

 im|)ortance. He would like to ask the author if the observed 

 temperatures agreed with the values calculated on the assump- 

 tion that the gases in the cylinder behaved as a perfect gas, and 



I'V 

 that -- was constant during the whole stroke. Differences 



l>etween the olwerved and calculated values might be due to dis- 

 sociation, and not entirely to lag in the thermometers. It was 

 astonishing that even the fine wires employed were able to follow 

 the rapidly varying temperature, and he would like to see some 

 special experiments made to test this point. Prof. Cap|X'r 

 showed a diagram gi^'ing the values calculated on the assump- 



PV 

 tion that — = constant. In such a calculation it was necessary 



to assume some tem|)erature as a starting-point, and in general 

 this tem|)erature was obtained from an analysis of the exhaust 

 gases, so that the calculated cur\e is most likely to depart from 

 the Inith at the commencement of the stroke. He, Prof. 

 Cap|x.T. ho]X'd that the author would be able to accurately 

 determine the tem|x;rature of some one point of the siroke, anil 

 he suggested! that the |x)int where the observed curve crossed 

 the theoretical curve would be the most suitable one for this pur- 

 pose. Such a point must exist, since at the commencement of 

 the .stroke the lag causes the obsen'cd temperature to be too low, 

 while at the end of the siroke the observe<l temperatures are too 

 high. Mr. ISurstall finds a curious bump in his curves, and it 

 is curious that a similar bump exists in the calculated curves. 

 From the constancy with which this bump appears, it would seem 

 lh,it it niusi have sfime physical meaning. It was important to re- 

 n)en)l)er that the expansion in the g-os-engine cylinder is not adia- 

 Intic, for heat is lK>lh abstracted and generated during the stroke. 

 Mr. lilakesley suggested that since the temperatures dealt with 

 were sufficient to make the wire red-hot, the question arose whether 

 lag might Ik: investigated by the wire l>eing examined by means 

 of Becquerel's phosphorosco]x;, at a known interval after the re- 

 moval of the s<iurce of heat. .Mr. (iriflilhs s.iid he considered 

 an im|x>rtant source of error was the large thermal capacity of 

 the leads when the working wire w.as so very short. He 

 ihought it would lie |K)ssihlc to standardise the thermometers 

 under conditions similar to those which occur in the engine 

 cylinder. Thus |>erhaps alternate gushes of air at o" and 

 ICX3' C. might l>e used. The use of golil to attach the fine wire 

 to the leads was objectionable, since the gold must permeate ihe 

 platinum for quite .an .appreciable fr.aclion of the whole length of 

 the wire. He would like to know whether the change in S 

 nlliukd to by the .author occurred with Ihe first explosion, or 

 r it was a gradual one. Mr. Knright |)ointed out Ihat 

 lire of the working substance in a gas engine varied 

 the stroke. Prof. I'erry said that the change in the 

 volume of the gases before and after combustion did not 



1 to more than I '25 [kt cent. Mr. K. Wilson said he 



Ihought il was most in)|y>rtanl to shorten the time of contact, 

 since at present Ihe galvanometer readings corresponded to ihc 

 mean temiKTature f>ver a range of abfiut 5 [xrrcent. of the whole 

 stroke. Il might Im; |X)Ssible lo make use of a condenser to get 

 over this difficully. Prof. Kllckcr said lhat Ihe Kew Observa- 

 tory were making nrrani;emenls lo undertake the testing of 



plnlinun 'i- ■ - Mr. Knrighl suggested lhat with 



a ver> .| currents might cause errors. 



Mr. Kh iind thai Ihe method of determining 



Ihc 7.ero |>oinl of Ihermomelcrs, by means of melting ice, was 

 far from salisfaclory, and thai the results obtained could not be 

 depended upon lo wilhin o''l C. The author, in his reply, said 



NO. 1338, VOL. 52] 



the only chemical action on the wires he thought proluble was 

 the formation of a carbide, .\fler several hours use, however, 

 the wires appear quite bright and clean. —Mr. N. V. Deerr 

 read a jxiper on the thermal constants of the elements. The 

 object of the paper is to establish the following laws: If T 

 denote the melting-point on the absolute scale. C the mean co- 

 efficient of exp-insion between ?ero and the meUing-poinl, S Ihe 

 mean specific heat, and L the latent heat of fusion, then, for any 

 family in Mendeleefs i)eriodic classification, the following 

 relations will hold between melals and metals, and between non- 

 metals and non-metals : 



(t+|)c = 



const. 



const. 



= const. 



TC 

 LC 



S 



In Ihe absence of other data, the mean values of C and .Sliclween 

 0° and 100° have been taken. .\noinaIous values are ob- 

 tained in the case of gold and mercury, if these melals are in- 

 cluded in their usual |«silions. The author considers thai Ihe 

 thermal constants indicate that gold ought to be ]ilaceil among 

 the transition elements. He further proposes to place mercury 

 in a new group to come before the lithium group. Such a group, 

 he suggests, would contain hydrogen, argon and mercury. The 

 ])aper concludes with an attempt to justify the expression 



V' 



S> 



-^ -^j C = const. 



on theoretical grounds. Dr. Gladstone considered that the paper 

 containc<l valuable numerical relationship, and lhat Ihe secomi 

 and thirtl formul.e were much more strongly sup|K)rled by Ihe 

 data given than Ihe first formula. He. the speaker, had 

 previously noticed lhat the elements of llie Iransilion group 

 might be subdivided into sub-groups, and lhat Ihe [elemenls ol 

 each of these sub-groups were particularly closely relaled. Ik- 

 agreed with the author that gold ought not to be included in the 

 first group. Mr. !•". H. Neville s;iid that since the author did 

 not give Ihe source of the ilala he had einployeil, most of Ihe 

 results given were rather indefinite. For example, while Ihe 

 author gives S70' as the melliug-poinl of aluminium, Mr. 

 lliiycock and himself had found the value 927". The value 

 of the latent heat of aluminium given was 29'3, while I'ionchon, 

 in a recent ]>a|)er in the Complts reiidiis, gives the value 80. 

 Theoretical considerations appear to indicate lhat 80 is 

 the minimum value possible. The author assumes that when 

 you heal a substance from the absolute zero lo its melting point, 

 all Ihe energ)- supplied is expended in Ihe work of expansion. 

 Some of Ihe heal, however, nuisl be employed in changing Ihe 

 kinetic energy of Ihe molecules, e\en in Ihe case of a solid. 

 Prof. Worlhington said that in some cases the amount of work 

 done against cohesive forces between 0° and 100° was much less 

 than one ten-thousandth of Ihe whole amount of energy su|)plied. 

 Mr. C'.riffithssaid he did not believe in any generalisalion which 

 dei>ended 011 the values of the specific heats deUrmined 

 between o" and 100°, the rale of change with leinperalure 

 of specific he.it being so great. The .author in his 

 reply said he had made every endeavour lo obtain 

 the most accurate data for his calculations. The value 29'3 

 for the latent heal of aluminium was obtained from a paper by 

 J. t'l. Richards. — A jiaper on an eleclroinagnelic effect, by Mr. 

 !•'. \V. Bowden, was postponed lill Ihc next meeting. 



Entomological Society, June 5. -The Kighl Hon. 

 Lord \Valsiiif;ham, I'. U.S., X'icePresideiil, in Ihe chair. 

 -Dr. Sharp, I'.K.S., exhibited, on behalf of Dr. ('.. D. Ilavi- 

 land, Iwci species of Cii/olrniii-i from lii>rneo, Ihe inilividuals 

 being alive and apparently in good heallh ; one of Ihc 

 two small communities (which were cimlained in glass 

 lubes) consisted of a few individuals of the immature sexual 

 forms and of a ncoleinic queen ; this latter had increased some- 

 what in size during Ihc eight moiilhs il had been in Dr. Havi- 

 land's possession, iml no eg^;s had been deposiled, neither had 

 any of ihe ininialure individuals developed inio winged forms. 

 The second lommunily exhibited lonsisleil entirely of the im- 

 mature sexual forms, and Ihiscommunily had produced numerous 

 winged adulls while il had been in Dr. Haviland's po.sses.sion. 

 .Specimens were also exhibited to illustrate the ncoleinic forms 

 Ihal were pr(«luced in Horneo after a communily had been 

 artificially orphaned. As regards these, Dr. Sharj) expressed 



