Jan, 9, 1890] 



NATURE 



225 



On seeing it in print, however, a natural answer occurs to me, 

 which it may be worth while to give. The whole point of the 

 reasoning depends on assumed properties of vacuum. 



The assumptions are as follow : — 



(i) That a perfect vacuum is an absolute non-conductor of 

 electricity. 



(2) That no contact EM.F. exists between a metal and a 

 vacuum. 



(3) That vacuum has a specific inductive capacity. 



Grant all these, and the argument is sound. Decline to 

 admit any of them, and it proves nothing. Break down the 

 first two of them, and it proves too much: it proves the non- 

 existence of any thermal contact-force whatever between con- 

 ductors. For if there were any E.M. F. at the metallic contact, 

 and none at the other or vacuum contacts, a continuous current 

 would flow, propelled by energy derived from a cold place. 



This argument is indeed the ordinary one to i:)rove that the 

 algebraic sum of the E.M.F.'s at all the junctions of a closed 

 conducting circuit in which no energy but heat is supplied must 

 be zero when the temperature is uniform. 



The proof scarcely holds when insulators are interposed, 

 though \}ci& fact may be true nevertheless. When chemically 

 active substances with their extraneous supply of energy are 

 interposed, the fact itself is no longer true. But how do we 

 know what is true when vacuum is interposed ? The hypothesis 

 on which the argument is founded is a baseless conjecture. 



But it may be said, Are not the hypotheses probable? Do 

 you not yourself believe them? I believe in (i) and (3) pro- 

 visionally, but certainly not in (2). The contact E.M. F. be- 

 tween two substances is probably some surface action or skin 

 phenomenon, and I see no reason why it should not occur as 

 well in the boundary between metal and void as in the boundary 

 between one metal and another. Indeed, it is not improbable 

 that the sum of the E.M.F.'s in every circuit of chemically inert 

 substances, whether conducting or not, and inclusive of vacuum, 

 is zero under uniform temperature conditions. 



All that is wanted to establish this is the knowledge that in a 

 circuit of any one substance at non-uniform temperature the 

 total E.M.F. shall be zero,^ or that the Thomson effects in a 

 single substance always balance each other ; i.e. that the total 

 E.M.F. in a circuit shall depend on a potential function of 

 temperature, ox d\i=f'{t)dt. 



Now it is quite true that this /'(/) is the Peltier coefficient 

 divided by absolute temperature, and that/(/) in its most general 

 form contains an arbitrary constant, but what of that ? Nothing 

 is known of f{t) except that it is a potential function : it is not 

 known to represent any physical effect. I never said that the 

 Peltier effect enabled us to find the most general form of the 

 function/!/) ; I said it gave us the E.M.F. at a junction. 



And there is much ground for the assertion ; for it is easy to 

 show that in a simple AB circuit, with junctions at /^ and t^, 

 the total E.M.F. is 



E 



n, 



n2 + /(e 



(0A - ®Adt ; 



just as if the resultant E.M.F. were the algebraic sum of two 

 Peltier E.M.F.'s and of two Thomson E.M.F.'s. 



My only contention is that this equation, which is undeniably 

 true when the IT are interpreted as heat-coefficients, is also true 

 and immediately interpretable when they stand for contact 

 E.M.l'.'s. The burden of proof as to the physical existence of an 

 unnecessary and in every sense arbitrary constant rests with 

 those who doubt this simple explanation. 



It is difficult to zee how a doubt can arise, or how the Peltier 

 and Thomson productions or destructions of heat can be ac- 

 counted for without local E.M.F.'s. Nohow, so Dr. Ilopkinson 

 has proved, and I also have insisted {F/iiI. Mag., October 1885, 

 .ind March 1886), except by some wildly gratuitous assumption 

 of an actual physical specific heat for electricity, dependent on the 

 temperature and on the metal in which it happens to be. 



Liverpool, December 14, 1889. Oliver J. Lodge. 



Mirages. 



The article in Nature of November 21, 1889 (p. 69), recalls 

 to me mirages I saw in March 1888, while travelling in the 

 I'^ast on the steam yacht Ceylon. 



On the 29th we were crossing the Black Sea from Sebastopol. 



' Hopkinson virtually pointed this cut, PhiL ]\fag., October 1885. 



It was a fine cool day and quite calm. In the afternoon a false 

 or mirage horizon about 3° above the true one was visible for a 

 few hours. No objects \vere within range of vision. The 

 mirage disappeared as the sun declined. 



The next day was very much warmer, and we saw a more 

 marked and interesting mirage in the afternoon as we were 

 steaming across the Sea of Marmora away from Constantinople. 

 In this case it appeared only in the west, and objects were seen 

 reflected in an inverted position. A small conical-shaped island 

 was seen with its inverted image at times distinct from and at 

 times blending with the original. The image was distinctly 

 seen of some land, which was actually below the horizon. The 

 mirage of the reflection of the sun in the sea was, when seen 

 through a glass, especially beautiful. It resembled a glorious 

 cataract of golden water. This mirage lasted till quite the dusk 

 of the evening, and then gradually thinned down and died 

 away. 



I do not know whether mirages at sea are uncommon ; but as 

 the officers on board did not remember seeing one before, I 

 thought these instances might be worth recording. 



Arthur E. Brown. 



Thought Cot, Brentwood, December 31, 1889. 



Self-luminous Clouds. 



I AM very sorry that I took no notes, some six or seven years 

 ago, on the first and only occasion of my seeing self-luminous 

 clouds, but though I can give neither date nor positions, the 

 following facts are still fresh in my memory. 



Passing through Bushey Park after dark, I noticed an aurora 

 borealis, and, as I had only recently seen the rather rare 

 phenomena of the rays of the setting sun converging towards a 

 point in the east, I followed the direction of one of the principal 

 lieams of light towards the south, when, at a point somewhat 

 south of my zenith, I noticed an equatorial belt of luminous 

 clouds. I found that each cloud belonged to a ray, and faded 

 and brightened with it, but was separated by about 60° of clear 

 sky. This belt of clouds extended down to the western horizon, 

 the eastern one was obstructed by trees, while shortly afterwards 

 small dark clouds appeared on that side, and the sky soon 

 became overcast. 



The luminous clouds were quite transparent, so that even 

 faint stars could be seen through them when at their brightest. 

 I have heard from Scandinavian captains that these luminous 

 belts are sometimes seen in northern latitudes, and are sure signs 

 of bad weather. I have written these few remarks in the hope 

 that those of your readers who may have the chance of seeing 

 an aurora borealis will also look out for these clouds, and if 

 possible determine their position. C. E. Stromeyer. 



Strawberry Hill, January 4. 



The Revised Terminology in Cryptogamic Botany. 



The anglicized forms of most of the terms in common use, 

 employed in the " Hand-book of Cryptogamic Botany " recently 

 issued by Mr. G. Murray and myself, have not up to the present 

 time found much support from our fellow-botanists. I propose, 

 therefore, to give, in some detail, the reasons which have 

 induced us to adopt them, and to urge their general use on writers 

 on cryptogamic botany. For this purpose we will take as our text 

 extracts from three reviews of the " Hand-book," marked, as all 

 the critiques have been, with only one or two exceptions, by a 

 generous appreciation of the difficulties of our task, and a too 

 great leniency to the many shortcomings of the work : — " The 

 most conspicuous, though not the most important, of these 

 [changes] is the adoption of anglicized terminations for Latin 

 and Greek technical words. This is a matter in which it is 

 hard to draw the line aright. ... As a matter of taste we 

 think the authors have gone much too far in this direction. 

 They complain of the ' awkwardness and uncouth form of these 

 words ' ; we should have thought the reproach applied much 

 more strongly to 'coenobe,' 'sclerote,' 'nemathece,' and 

 'columel'" (Nature). "An Englishman may guess what 

 ' archegone ' is short for, for example ; but why puzzle a 

 foreigner with a new form of a word with which he is familiar 

 in every treatise hitherto written on the special subject in any 

 European language?" (Academy). "Too sanguine expectations 

 on this head might well be toned down by remembering the 

 complete failure of the somewhat similar experiment made by 

 Lindley. . . .Primworts, spurgeworts, bean-capers, and hip- 



