March 1, 1893.] 



KNOW J^ EDGE 



49 



explosive mixtures tested in our laboratories. But there is 

 here no necessary analogy. In the explosive niixtures of 

 our laboratories the propagation of chemical action is 

 relatively slow, because each successive part of the mixture 

 must needs be ignited and exploded by the heat evolved 

 before the action can pass onward. In the explosive 

 mixtures of the solar atmosphere such a heating can- 

 not take place. C/ieiiiicnl action is not caused there bij a 

 ciinsiiUraliie luatin;!, hut hij tin almost infinitesimal coolimj. 

 The cooling required is almost infinitesimal because the 

 rate of dissociation of solar matter necessarily corre- 

 sponds with the temperature. The cooling required 

 will occur, and cross the solar atmosphere from the 

 bottom to the top with the velocity of radiant heat, 

 when the heat radiated by the photosphere is suddenly 

 diminished by the forming of a spot or pore. If that 

 sudden cause of cooling is not immediately attended by 

 chemical action propagated also with the velocity of 

 radiant heat, it is because the dissociated matter is not 

 always ready to combine on the least loss of heat. 



Another objection is " that the rapid translation towards 

 or from the eye of the luminous condition in gases should 

 give rise to greatly widened lines, unless the kindling 

 action was instantaneously followed by extinction." Here 

 I must remark that the prominence lines are sometimes 

 greatly widened, and that in the cases where displacement 

 is observed without widening there is no great difficulty in 

 conceiving that the kindling is instantaneously followed by 

 extinction 



The last objection of Miss Gierke is that the forms of 

 nianij prominences seem to indicate real movements of 

 matter ; but I think that such indications must be accepted 

 with caution. The cirri-clouds of our own atmosphere, for 

 instance, seem also very often to be formed by real move- 

 ments of their cloudy constituents. But that apjjcarance is 

 surehj misleadini/ when it is shown by cirri suddenly 

 appearing, and then covering almost instantaneousli/ a great 

 part of the heavens with their straight or elegantly curved, 

 parallel or divergent filaments. Many prominences have, 

 moreover, very capricious and frequently broken forms, 

 which show often-changing entirely discordant directions 

 when suddenly appearing in the liigher layers of the solar 

 atmosphere, running in different directions with irregu- 

 larly varying and speedily increasimj velocities. They do 

 not, to my mind, convey any idea of actual motion in a 

 rcsistinij medium. If the prominences, as we see them, 

 consist of several prominences superposed on one another, 

 the difficulty of explaining their sudden outbreaks in the 

 higher layers of the solar atmosphere by the hypothesis of 

 upward projection increases ; for if it is difficult to 

 assume that one prominence should have its connecting 

 stem with the underlying photosphere hidden by some cool 

 mass, the difficulty is increased when there are several 

 co-existing prominences. 



Your last objection is that solar observations " do not 

 seem to point to an undisturbed stratification of the solar 

 atmosphere in which the heavier gaseous compounds sink 

 to the bottom." According to you the matter in the solar 

 atmosphere is well churned and evenly mixed, and its 

 different spectra at different levels are caused by differences 

 of temperature. Your explanation, however, does not 

 explain why the matter indicated by the spectroscope in 

 the lower regions is generally much header than the matter 

 above. There is no reason why metals should require a 

 higher temperature to show lines than hydrogen. It seems 

 also to me very difficult to comprehend that in the close 

 neighbourhood of the immense photosphere the temperature 

 shciuld be so low that iron vapour would not show its lines, 

 for we know that even in the Bessemer flame those lines 



are easily detected, and that they have also been observed 

 in the immensely rarefied matter of comets when heated in 



the proximity of the sun 



Yours faithfully, 



Dr. A. BliESTER, .Jz. 



Delft, Holland, February IGth, 1893. 



[I have printed the parts of Dr. Brester's reply which 

 seemed to me most cogent. His first point seems to 

 amount to the statement that we know nothing about the 

 chemistry of hot bodies, and therefore explosions on the 

 sun mai/ be propagated with the velocities which his theory 

 requires. They are certainly not propagated in the sun 

 with the velocity of radiant heat, and if it were possible 

 that a spot or pore in the photosphere could throw a 

 shadow, as Dr. IBrester suggests, the cooling effect would 

 be produced in radial lines from the sun, and not in the 

 curiously contorted forms which the prominences present. 

 On the one hand we have vague guesses at what the 

 velocity of explosions with an unknown chemistry might 

 be, and on the other we have the fact that the observed 

 upward velocity of several of the great projection promi- 

 nences has been shown to correspond with the velocity of 

 matter projected upwards under the influence of solar 

 gravity in a thin resisting medium. 



How Dr. Brester reconciles his theory of explosions with 

 a quiescent solar atmosphere I do not understand, for 

 explosions seem to me to imply disturbance and the actual 

 displacement of matter, so that if there were no such thing 

 as gaseous diffusion the vaporous material of the sun 

 would be speedily mixed. The spectra of terrestrial 

 elements which can be recognized in the sun are not 

 arranged vertically in the order of their vapour density, 

 and we know that in the spectra of metals in the electric 

 arc some lines are longer than others, that is, some of the 

 lines are only visible in the hot central region of the arc, 

 while others extend into the cooler outer region. — A. C. 

 Kanyaed.1 



THE LIFE OF STARS. 

 To the Editor of Knowledge. 



SiH, — I am not aware that the theorists who have 

 written on the evolution of stars have hitherto attempted 

 to trace out the effects of the supposed changes on the 

 stellar spectra. I venture to ofier a few suggestions on 

 this topic. 



Starting with the nebular theory, some form of which 

 seems to be generally accepted, we would have a bright- 

 line spectrum to commence with. If the nebula was 

 distant and had shrunk to small dimensions it might be 

 otherwise undistinguishable from a star. The continuous 

 spectrum would not appear until condensation (probably 

 into the liquid form) had commenced. But this might 

 commence in two ways— the formation of a central solid 

 (or rather liquid) nucleus and the formation of luminous 

 clouds. The sun's photosphere probably is of the latter 

 character. The solar spots seem to be breaks in the 

 clouds, revealing glimpses either of a solid nucleus or of a 

 second cloud-bank below. But these breaks are caused 

 by disturbances which somewhat complicate the pheno- 

 mena. 



The light of a central nucleus would reach us through a 

 great depth of absorbing gases. The dark lines would 

 therefore in all probability be numerous and strongly 

 maiked ; but bright lines might, I think, also be expected 

 to appear. The hemisphere of the star turned towards us 

 would consist of a solid portion seen through the gases, 

 and another portion (and at early stages a much larger 

 one) purely gaseous. The gas-light would thus come to 



