NATURE 



[January i6, 1896 



a large long mass of sand, suddenly giving way and falling for- 

 ward from the vertical face of a sand -bank or river- bank on to 

 the water, will produce, and the very long distance it will be 

 heard up and down the river. 



In the vicinity of such great rivers, at such a spot as 

 Chilmari Ghat, which I know well, we must be rather sceptical 

 with regard to sounds being any other than of this nature. The 

 reports are very likely to be repeated, for the wave produced on 

 the water will set another mass falling some way off, and even 

 the wave of air may do the same to a mass just tottering, and 

 i)rlng it all down together (see also Report in Proceedings of 

 Asiatic Society, Bengal, p. 207, regarding the echo from one 

 side of the river to the other, so well described by Mr. Manson). 



Mr. W. T. Blanford has suggested to me landslips for the 

 reports heard near Buxa ; but this is a sound I have heard on the 

 bursting of a landslip dam in the Himalayas : the sound was of 

 the nature of a rumble, not a report. A fall of rocks into a 

 valley from any height would produce, I think, a still more 

 prolonged rumbling sound— more like that of the avalanche. 



Shalford Park, January 7. H. H. Godwin-Austen. 



Allow me to mention, for the information of those of your 

 readers who take an interest in this subject — a considerable 

 number, to judge from the correspondence that has lately 

 appeared in these columns — that a very complete account of the 

 phenomenon is now being published in del et Terra, a review- 

 devoted to astronomy, meteorology, and terrestrial physics, 

 published twice monthly at Brussels, from the pen of M. Ernest 

 Van den Broeck, curator of the Royal Museum of Natural 

 History of Belgium, who has devoted much time and labour to 

 a study of the subject. The first article, containing a complete 

 historical review of the subject, appeared on December i ; the 

 second, on the i6th, dealt with the accounts of the phenomenon 

 received by M. Van den Broeck from various observers in 

 Belgium ; and future articles will be devoted to a study of the 

 facts as reported from Bengal, to the causes of the phenomenon, 

 and to the methods employed in determining its origin. 



I may mention that the true rendering of the word mist- 

 poeffers, the name given to the sounds by the fishermen, is not 

 fog-dissipators, but, as appears from a note on the etymology of 

 the word by one of M. Van den Broeck's correspondents, it 

 should more correctly be translated fog-belchings ox fog-hiccups, 

 the French rendering of the word poeff being renvoi or hoqitet. 

 The point is of some importance, since the sounds do not 

 appear to have any effect, as one might imagine from the 

 signification of the name, as it first appeared in Nature, in the 

 dispersal of fogs. T. D. La Touche. 



Boring a Coral Reef. 



Will you be good enough to allow me to correct a somewhat 

 important oversight in the note in last week's Nature, on 

 the work of the Coral Reef Committee of the Royal Society. 

 You mention that the Royal Society has granted the sum of 

 ;^8oo, and the Government a gunboat, for the expedition Even 

 ■with this help, however, it would have been impossible to under- 

 take the work unless the Department of Mines of the New South 

 Wales Government, largely through the good offices of Prof. 

 Anderson Stuart, had granted to the Committee the use of a 

 complete set of boring tools and appliances, with an engine and 

 boiler. Further, the Department has relaxed the somewhat 

 stringent restrictions usually placed on the use of the tools, owing 

 to the difficulties which will be experienced in a waterless and 

 sparsely inhabited island. For such generous assistance, and to 

 Prof. Stuart, who has helped in many different ways, the Com- 

 mittee cannot be too grateful. W. W. Watts, 

 (Joint-Secretary of the Royal Society's Coral Committee)! 



Sutton, Surrey, January 11. 



Variability of Red Stars. 



AccoRDiNc; to a notice, which I found in Nature's " Astro- 

 nomical Column" of November 14, 1895, my theory of the 

 variability of red stars should demand, that the bright lines of 

 hydrogen should be produced there at a low temperature and in 

 a condition never tested by experiments. 



I will try to show that this objection is not valid. 



The bright lines in red stars are considered by me as caused 

 by the recently much studied phenomenon of luminescence, i.e. 



NO. 1368, VOL. 53] 



by a production of light not due to elevation of temperature, an d 

 thence not subjected to Kirchhoff^s law of absorption. 



In the cooled atmospheres of the red stars two causes of 

 luminescence can be expected— two causes, which according to 

 the relatively low temperature of the combinable elements and 

 condensable vapours, with which these atmospheres are filled 

 up, must act there much more energetically than in other stars. 

 These two well-studied causes of luminescence are chemical 

 action and electricity. 



If we first consider chemical action, it may be stated, by the 

 way, that the supposed chemical combination in a cooling star 

 does not necessacily require a very low temperature. Acetylene, 

 for instance, can be generated in the electric arc. Now we know 

 that acetylene is a compound actually present in the atmospheres 

 of some of the most cooled stars. It is abundantly absorbing 

 in the stars \\\ h and, according to Prof. Lockyer, not unlikely 

 radiating in the stars of his Group II. (Vogel's Class III a). 



If therefore we assume that it is principally acetylene, which is 

 generated in the cooling atmospheres of the red stars, bright lines 

 of hydrogen may be expected there. These lines are then 

 precisely caused as those of sodium, potassium, lithium and 

 thallium in the experiments of Pringsheim, being absent as long 

 as the sodium vapour is only strongly heated, but appearing 

 immediately as soon as in this heated vapour some chemical 

 change occurs. 



It must be conceded, however, that in the case of hydrogen 

 this chemical luminescence has not yet been actually observed 

 in our laboratories. But in the case of hydrogen we have 

 another luminescence, which is daily experienced. It is the 

 glow in Geissler tubes, where (whatever may be the heat of the 

 individual shining molecules) the average temperature of the 

 glowing gas does not necessarily exceed the common temperature 

 of our atmosphere, and cannot therefore be considered as the 

 cause of the glowing of the gas. 



As this luminescence is caused by electric discharges, and as 

 such discharges (according to what we know about the electrical 

 phenomena connected with the formation of clouds and hail in 

 our own atmosphere) are likely to be expected in atmosplieres, 

 which are filled up with vapourous matter ready to condense in 

 clouds, the hypothesis of an electrical luminescence of the 

 hydrogen in the atmospheres of red stars seems very plausible. 

 That hypothesis demands that the bright lines should be 

 especially conspicuous in stellar atmospheres, where the alterna- 

 tion of vapourisation and recondensation is also very conspicuous, 

 i.e. (according to my theory explained in the November and 

 December numbers of Knowledge) in the atmospheres of red 

 variables and Nov?e, where that intermitting condensation and 

 vapourisation of dark obscuring cloudy matter is the very cause 

 of the variability. Now we know that this demand is fully 

 verified. Prof E. C. Pickering has stated as a rule that, with 

 perhaps a single exception, every red star with bright hydrogen 

 lines is eo ipso variable. And this rule has proved to be so sure, 

 that Mrs. Fleming could discover numerous new variables from 

 the bright lines of their spectra {Astrophys. Journal, I. p. 27, 

 411 ; II. p. 198). 



Eventual bright lines in the stars of Vogel's Class I. and II. 

 may be explained perhaps (as Prof. Scheiner has recently shown 

 in his " Untersuchungen liber Spectra der Helleren Sterne," p. 

 223) by the hypothetical presence of a gigantic incandescent 

 atmosphere, whose radiation around the bundle of rays coming 

 to us from the star's much smaller photosphere is greater than 

 the absorption it causes in that bundle ; but in the more cooled 

 atmospheres of red stars the chemical compounds there present, 

 attest a temperature relatively so low that we cannot conceive 

 that atmospheres to be filled up to such a gigantic height with 

 hydrogen so enormously heated as to become (if that still 

 doubtful phenomenon is possible) bright shining by incan- 

 descence. 



Such enormously heated hydrogen and chemical compounds 

 cannot permanently coexist in a stellar atmosphere. The 

 coexistence of chemical compounds and bright shining hydrogen 

 is only possible if the brightness of the latter is due to 

 luminescence. 



The ideas here suggested may be considered as an instance of 

 a likely fruitful application of the study of luminescence to 

 stellar spectroscopy. If they are right, they give, I think, a 

 plausible explanation (i) of the connection between the 

 variability of a star and the frequent brightness of its spectral 

 Hnes, and (2) of the very remarkable fact that both variability and 

 bright lines are so often observed precisely in those stars, whose 



