198 



NATURE 



\yan. 2, 1890 



surface is more nearly black than white, which seems to render 

 the existence of snow fields upon it less probable, unless they are 

 covered with voloanic dust, as the end of a glacier usually is 

 with rock debris. 



But even if we take Mr. Reade's view, it is still conceivable 

 that steam may have been the explosive agent in the moon's 

 volcanoes, while her internal temperature was very high, and 

 that the resulting water may have been subsequently absorbed 

 after the body became cool, because the water would occupy 

 less space within the interstices, which this theory of imbibition 

 postulates, than the equivalent vapour did, when the temperature 

 was high. The case of the earth would not be a parallel one, 

 because it has not yet cooled. 



Although not myself a selenologist, I have a suspicion that very 

 little is known about the constitution of the moon ; and that it is 

 not even certain that its enormous craters are all of them really 

 volcanic. It has been admitted by Prof. Darwin, in discussing 

 the subject with Mr. Nolan, that on his view of the genesis of the 

 moon it must have originally existed as a " flock of meteorites." 

 These falling in during the later stages of the building up of its 

 mass would have produced pits on a viscous surface, much like 

 some of the craters. 



At any rate it seems unsafe to rely upon arguments respecting 

 the condition of the earth's interior, of which we know little, 

 drawn from that of the moon's body, of which we know less. 



Harlton, Cambridge. O. FiSHER, 



Exact Thermometry. 



The interesting experiments of Dr. Sydney Young, recorded 

 in Nature of December 19 (p. 152), seem to leave no doubt 

 that the main part of the permanent ascent of the zero-point of 

 a mecurial thermometer, after prolonged heating to a high tem- 

 perature, is not due to compression of the bulb — rendered more 

 plastic by the high temperature — by the external atmospheric 

 pressure. Researches on the effects of stress on the physical 

 properties of matter have convinced me that the molecules, not 

 only of glass, but of all solids which have been heated to a 

 temperature at all near their melting-point, are, immediately 

 after cooling, in a state of constraint, and that this state can 

 be more or less abolished by repeatedly heating the solid to 

 a temperature not exceeding a certain limit, and then allowing 

 it to cool again (it is not only the heating but the cooling also 

 that is efficacious). It appears that the shifting backwards and 

 forwards of the molecules, produced by this treatment, enables 

 them to settle more readily into positions in which the elasticity 

 is greatest and the potential energy is least. 



This "accommodation" of the molecules, as Prof. G. 

 Wiedemann and others have called it, is, as one might suppose, 

 attended with alterations of the dimensions and other physical 

 properties of solids, and is not confined to the release of mole- 

 cular strain set up by thermal stress, but is extended to the 

 strain set up by any stress whatever. As years roll on, the 

 time of vibration of a metal pendulum gradually alters (and so, 

 no doubt, do the lengths of our standard measures), the bulb of 

 a thermometer diminishes in volume, a steel magnet parts with 

 more or less of its magnetism, a coil of German-silver wire gains 

 in electrical conductivity, &c. The changes in all these cases 

 would probably be far less than they actually are if the tempera- 

 ture throughout the whole time could be maintained constant ; 

 but this last is not the case — heating and cooling goes on more 

 or less every day. We may assist the effect of time by artificially 

 increasing the range of temperature, but it would appear that 

 we must not exceed a certain limit of temperature, which limit 

 depends partly upon the nature of the substance and partly upon 

 the stresses that are acting upon it at the time. Thus, the in- 

 ternal friction of a torsionally oscillating iron wire which has been 

 previously well annealed may be enormously diminished by 

 repeatedly raising the temperature to 100° €., keeping it there 

 for several hours, and then allowing it to fall again. The amount 

 of diminution of internal friction depends upon the nature of the 

 wire, and on the load which there is at the end of it (if the load 

 exceeds a certain amount, the friction is increased instead of 

 diminished). In attempting to "accommodate " the molecules 

 in this manner the heating must, at any rate in some cases, be 

 prolonged for several hours, and the substance should then be 

 allowed to remain cold for a still longer period. 



I have not had much experience with glass, but I think it prob- 



able that the settling down of the zero-point of an ordinary ther- 

 mometer into its ultimate position could be very materiall)' 

 facilitated by the heating and cooling process mentioned above. 



Herbert Tomlinson. 

 36 Burghley Road, Highgate Road, 

 December 2^, 1880. 



Self-luminous Clouds. 



Without venturing to call in question the occasional occur- 

 rence of self-luminous clouds, I may be permitted to relate an 

 observation which seems to reveal a possible source of error in 

 the records of such phenomena. 



On June 14, 1887, about 10.45 p.m., I witnessed an ap- 

 pearance over the north-north-west horizon which struck me 

 as very remarkable. Amidst the strong glow of twilight a few 

 fragments of cirrus cloud shone with a pure white light having so 

 much the character of phosphorescence that it was difficult to 

 believe the objects were not self-luminous. Looking out again 

 an hour later, I found no trace of bright clouds, but in their 

 place were small bands of cirrus showing dark and grey against 

 the feeble twilight that remained. I could not but conclude that 

 the clouds in both instances were the same or similar, lit up by 

 the direct rays of the sun at the time of the first observation, and 

 having lost his rays at the time of the second observation. Had 

 they been self-luminous they should have become brighter 

 instead of darker as the twilight faded. 



It has been suggested to me that the bright clouds seen at 

 10.45 P-in. may have owed their brightness, not to the sun's rays 

 falling on them at the time, but to a temporary phosphorescence, 

 the result of exposure to the sun's rays in the day-time, and that 

 this temporary quality had died out in the interval between the 

 two observations. 



I think this explanation is unnecessary for the following 

 reasons. In the first place, it is certain that if a cirrus cloud 

 were present in the atmosphere at a sufficient height to catch the 

 sun's rays at 10.45 P-m. of a midsummer day, it would appear 

 as a bright object amidst the surrounding gloom. And, secondly, 

 there can be nothing incredible in the presence of a cirrus cloud at 

 that height, when the persistence of twilight proves the presence 

 of atmospheric particles of some kind at a greater elevation 

 still. George F. Burder. 



Clifton, December 19, 1889. 



Duchayla's Proof. 



I HAVE read with much interest the new proof given by Mr. 

 W. E. Johnson of " the parallelogram offerees," in Nature of 

 December 19 (p. 153), and regard it as deserving a place among 

 the best proofs that have been given. 



I think, however, that, in his criticism of Duchayla's proof, ISIr. 

 Johnson runs to excess, when he says, " To base the funda- 

 mental principle of the equilibrium of a particle upon the 

 transmissibility of force, and thus to introduce the conception of 

 a rigid body, is certainly the reverse of logical procedure." 



Duchayla's proof only requires us to suppose the transmission 

 of force by strings. A particle is unthinkable. In presenting 

 to a learner the conception of three equilibrating forces acting 

 on a particle, we cannot do better than represent the forces 

 by pulls in strings, and the particle itself by the knot where the 

 three strings are tied together. All the steps of Duchayla's 

 demonstration that the resultant force is directed along the 

 diagonal of the parallelogram can be presented in tangible form 

 with the aid of strings. I do not think this is an illogical or 

 unnatural procedure. J. D. Everett. 



Belfast, December 23, 1889. 



The Satellite of Algol. 



The results of Vogel's photographs as to the satellite of Algol 

 are of great interest to your astronomical readers. The ob- 

 servations made at Greenwich tended to the same result, but were 

 unfortunately intermitted before anything approaching certainty 

 was arrived at. 



Regarding it as certain that the variations of Algol are due to 

 the interposition of a satellite, the question of the slight change 



