30 



NATURE 



[May q. 1S95 



encounter were not independent, and our assumptions (Aon-rtwr 

 imprt>babU) would lie therefore entirely tiasol on our previous 

 experience with the direct inotion. Without such assumptions 

 we should have inferred, by the ordinary laws of probability, 

 that H would be likely to decrease. This is what I intended to 

 imply in my previous letter : but as I had use<l accented and un- 

 accented letters in my statement, I failed to make my nieaning 

 clear to Mr. Culverwell, who evidently found it ditticull to under- 

 stand a pnxif involving their use. 1;. M. Bryan. 



The Unit of Heat. 



I \v.\s glad to read Prof. lolys conmiunication in your is.sue 

 of May 2. for I have made many efforts to call attention to the 

 unsatisfactory nature of our present system of calorimetric 

 measurements, and now that a more powerful voice than mine 

 has been raised in favour of a change, I have some hopes of 

 progress. 



The iniliflerencc with which, as it appears to me, our physicists 

 regard this matter is probably due to several causes. They 

 ignore the fact that the science of calorimctry has recently made 

 great strides, and that an ambiguity as to the unit, which 

 fonnerly was of little consequence, has now become almost the only 

 bar to further pnigress ; also, as Prof. Joly has (winted out, our 

 system of calorimetric measurements has l>een .so weilded to 

 the method of mixtures, that the union has (wrongly) come to be 

 regarded as es.sential. 



.\s to Prof Joly's proposal, there is much to be said in its 

 fiivour. It is practical and definite. -At the same time the change 

 would be so radical, that I should not feel justifie<l in counting 

 myself as his disciple in this matter without serious consi<leration. 



My own inclination is rather in the direction of a (".d.S., or 

 alisolute unit, and the course adopte<l by Prof. Schuster and .Mr. 

 (iannon, in entitling their recent important communication to the 

 koj-af Society "The Specific Heat of Water," rather than the 

 ■■ Mechanical K(juiv.ilent of Heat," shows that a step has 

 already lieen taken in this direction. 



When we reflect on the attention and the labour which have 

 lieen devoted to the establishment of our present system of 

 electrical units, it is a cause for wonder that so im|>ortant a unit 

 .Ts that of heat .should have been left ill-defined and unregarde<l. 



I would propose that at the forthcoming meeting of the British 

 .Association, the attention of Section .\ should lie [wrticularly 

 directefl to this matter ; and it would prepare the way for such 

 action if those who have definite proposals to make wouUl, in the 

 meantime, communicate them to your columns. 



Cambridge. K. H. Ckikhiiis. 



j reverse it, sujierimposing the two ends as before, and sketch it 

 I in alongsiile his first curve (easily done by n^eans of oil-|iaiier), 

 then, if they iliffered, draw a fresh curve midway between the 

 two ; subsequently re-marking his examination |)a|>ers from this 

 smoothetl mean cur\'e ? .-\n illustration may be of use : let it be 

 founded on Fig. i , as it contains the less smooth curve. The 

 dark line is that of the marks first adjudged ; the light line, 

 the same curve reversed ; and the dotted line, the smoothed mean 

 curve of the two from which his ixq^ers are finally marked. 



Granting that the plus variations and the minus variations on 



the two sides of the mean nearly balance, the question would 



. appear to be — Would one lie justified in smoothing them in 



accordance with the generalisetl results of many such series ? 



I It involves some forcing of the examiner's marking into the 



j general mould, but would this Ix" n>ore than sufficient to correct 



Keferri.nc. to Dr. Joly's letter la.stweck, would it not be well 

 ilefinitely to adopt the "joule" as the only fundamental unit of 

 heat, an<l to realise distmctly that researches such .is those of 

 Mr. (jriffilhs. Prof. Rowland, and Ur. Joly are determinations 

 of the specific heat of water and of the latent heat of steam in 

 terms of^ it ? Oliver J. LniicK. 



The Examination Curve. 



The extremely interesting article, by Prof. I.loyd Morgan 

 (vol. li. pp. 617 619), on the graphic represenlatiim of the marks 

 given in an examinalirm, and of their great use to an examiner, 

 leads me to a.sk whether even this methiHl may not Ik; ileveloped 

 further with .idvantage to all concerneil, for, as Lloyd Morgan 

 says — " If, after an extensive set of i>a|x;rs has been looked over 

 an<l carefully marked, an interval <if time \k alloweil to elapse, 

 .inrl then the papers are gone over .-igain, the result of this re- 

 examination IS that the head ami tail remain practically 

 unchange<l, but thai there is not a little redistribution among the 

 moliocritics." In other words, the jx-rsonal eipialion of the 

 examiner varies, showing itself mostly in the miclille of the 

 curve. 



7 h.. hr.i ii,,,,,. i.,stri|(t. me on looking at Fig 2 (vol. li. p. 618), 

 " irity of the I wo halves of the curves, and on Irac- 



II 'ling thctracing half roundMi that theup|K-rend 



of ilit ir.ne-i lurve Ijccame superimposed U|xm the lower end of 

 th<- ori(Hn.il. and vur jfrsii, (he similarity was so marked as to 

 II '-I'l a larger numlnrr of ]iai)crs liccn ex- 



marked as the first set, the traced curve 

 w-....-. . ; ... other. 



If -' i-^'. why should not the examiner, after plot- 



ling III- 'hinks licsl, make a Irarini; "f this ■ iirM-. then 



NO, 1332, VOL. 52] 





27 2S 23 21 19 17 IS 13 II S 7 S 



Kxamiiut'i. Fii.. !. 



his |X"rsonal equation? On the other hand, the twn halves 

 — say from paucity of examiners — might be so dissimilar, 

 that the mean curve would iliffer very much from the original 

 form. In this ca-se, would it be possible to give any general 

 rule whereby one coultl be guided whether to adopt the mean 

 curve, or to remain satisfied with the original marks given ? 



In HerlxTt S|iencer"s " Principles of Sociology," (vol. i. p. 88) 

 are many references to the fact that "the children of .\ustra 

 lians, of Negroes in the Uniteil .States, of Ni'groes on the Nile, 

 of .\ndamanese, of New Zealanders, of .Sandwich Islanders [anil 

 others], are quicker than Kuropean children in acquiring simple 

 ideas, but presently slop short from inability to grasp the com- 

 plex ideas rea<lily grasped by K,uro|)ean children, when they 

 arrive at them." K. How.vRU Coi.i.iNs. 



.•\pril 29. 



Teaching Young Pheasants to Peck. 



It may interest Prof. Lloyd Morgan and others to know thai 

 when .Asamese find newly halcheil chicks in ihe jungles, they 

 have a s> stem of tciching the little ones to peck and pick up food, 

 without which, I am told, many of them would die. 



Walking down a ro.id one morning with a neighbour, we sud- 

 denly n<iliced a litlle ball of fluff between my feet, and I could 

 hardly avoid stepping on il, .is il stuck close to me ; almost 

 immediately another ap|KMred at my friend's feet, and we saw 

 they were newly-halche<l pheasants, the mother ))rol);ibly carrieil 

 off by some wilil cat. 



As it was difficult lo walk with*the.se little things running so 

 close and in the way, we lifted them into the short grass along- 

 side, and hurried on s<mie fifty yards. 



On returning we had forgotten them, but one ran out, and so 

 |X'rlinaciously stuck to my boots, that to save it I put il into my 

 jiocket, and on our arrival al the bungalow tried to feed it with 

 small fragments of hard-lioiled egg, rice, and white ants. Of all 

 these it look no notice. 



Next nmrning ihe olher chick w.is found al the fool of llie 

 Iningalow steps, having proliably followe<l us unnoiiceil the ilay 

 liefore. I then called my " Halm," as I could not gel them lo 

 eat. and he said "they must be //7//^A/." 



lie pill the gau/e wire cover they were under, and the crushed 



