Oct. 5, 1882] 



NATURE 



55: 



vivid idea of how rapidly the power of the air to hold 

 moisture in invisible suspension increases with the tem- 

 perature ; and remember that it is not until the quantity 

 of watery vapour accumulates to a still greater extent 

 than what air of such temperature can assimilate, that 

 there is spare material enough for producing rainfall. 

 Hence, while in Scotland a rain-band of intensity marked 

 2 usually produces a little rain, and 3 produces much, yet 

 in Lisbon during the same months the so-called rain- 

 band, but really only water-vapour band, may mark 4, 

 and yet no rain fall. But with 5 or 6, the temperature 

 remaining the same, down rain will come even in that 

 usually arid country. 



Again, whatever number of supernumerary observa- 

 tions any person may take, when his enthusiasm-fit is 

 upon him, he should never neglect his usual, regular 

 observation at a fixed hour, say 9 a.m. For if the wet- 

 bulb depression goes through a diurnal rise and fall 

 according to the hour, something of the same kind may 

 be looked for in the strength of the spectral water-vapour 

 band ; though fortunately it is not so very marked a 



feature there, because the upper strata of the atmosphere 

 are more constant in their composition from hour to hour, 

 than its lower beds in contact with earth and water. 



But why should I go on wearying readers of Nature 

 with these little details, when they can far better find out 

 such things for themselves, and often realise improve- 

 ments therein. See how well Mr. Rand Capron has 

 mastered the subject, in his " Plea for the Rain-band" in 

 Mr. Symon's Meteorological Magazine. How acutely 

 Colonel Donnelly appears to have detected in the second 

 water-vapour band of the spectrum, viz. that near the 

 solar C line (a darker part of the spectrum than that 

 occupied by the band near D, and therefore more difficult 

 to observe), an indicator of a different order of precipita- 

 tion from the atmosphere than ordinary rainfall. And 

 again I trust to be excused for mentioning here that my 

 friend, Mr. T. Glazebrook Rylands, has now accumulated 

 an immense deal of experience as to the advantage of 

 supplementing spectroscopic rain-band observations with 

 a polariscope equally portable. 



At present, when experimenting for further advance, I 



SCALE OF WAVE NUMBER PLACE. PER BRITISH INCH 

 lOAOO 42.S00 42.000 43.700 



SOLAR LINES 

 BY NAME OF 

 MATERIAL. FROM ANGSTROM 



The Water-Vapour band on the Red side of D 1 and D* 

 Edinburgh, through the average ol August 1882, at ic 



seen in the faintly illuminated N'orth-Western sky at 5° above the horizon, from Royal Terrace, 

 each morning, with a powerful spectroscope. Temp. = 6V'o, depression of wet bulb = 3° 'o. 



SCALE OF WAVE NUMBER PLACE. PER BRITISH INCH 

 K.400 42.500 42.600 4Z.700 



xtr 



SOLAR LINES 

 BY NAME OF 

 MATERIAL, FROM ANGSTROM 



1 September 4, 1882, on the eve of a whole week of very dry weather ; temp. = 55*'°> depression of wet bulb =5° 



rather prefer the spectroscope alone, but of greatly 

 increased size and power ; and it was not until very lately 

 that I fully experienced what can be done in this way 

 upon merely the faint light of the sky near the northern 

 horizon, a region seldom seen here without more or 

 less clouds and much manufacturer's coal-smoke, yet 

 forming a better constant of daylight than if I had 

 attempted to look southward into the neighbourhood of 

 the sun. 



On direct sunlight, whenever it can so very rarely in 

 this country be enjoyed, of course almost any spectroscope 

 will show multitudes of lines, and even split up telluric 

 bands into many fine lines ; but to see a large spectro- 

 scope accomplish nearly the same fact on merely low sky 

 illumination, gave me a new idea of the discriminating 

 powers of this marvellous modern apparatus, and im- 

 pressed me with the positive duty of trying to use it 

 quantitatively, as well as qualitatively. I append, there- 

 fore, a map of the lines and bands of the chief " rain- 

 band," so called, of the ordinary spectroscope, but now 

 as seen in the larger instrument through the average of 



last August ; and again, for a contrast, as it was seen on 

 one particular day, September 4, when a week of the 

 driest and coldest weather of the season was about to 

 begin. 



The hygrometer readings taken elsewhere conformed 

 pretty well to these descriptions ; but in their whole 

 variations from 2° or f for the earlier, and 6° for the 

 latter time, there was nothing to call up such intense 

 interest as the spectroscope's astounding fact of the 

 almost entire sweeping away on September 4 of the many 

 and rich details of the previous month, in so far, ot 

 course, as they were water-vapour spectral details. 

 Nature herself does therefore offer in the way of ground- 

 work for rainfall forecasting in the spectroscope, so large 

 an amount of material, that I do trust no one will under- 

 value it, until they have had practice, with an equally 

 powerful instrument with that I have just alluded to. Its 

 main features are, that the object-glasses of both colli- 

 mator and telescope are 2-25 inches in diameter ; each of 

 its two prisms is 7 inches long and 3'5 inches square at 

 the end, and contains bisulphide of carbon at a refracting 



