Z. Bell — Rainband Spectroscopy. 347 



Akt. XLY. — Rairiband Spectroscopy ; by Louis Bell. 



It is now more than ten years since Professor Piazzi Smyth, 

 pointed out that the absorption bands of aqueous vapor in the 

 solar spectrum were likely to be of considerable service in 

 meteorology ; but although his observations were evidently 

 successful and have been often repeated with equally good 

 results, little progress has been made in the practical appli- 

 cation of the principle. This lack of results appears to be 

 due to two causes. First, as the method seemed to offer an 

 easy way of predicting the weather, it at once fell into the 

 hands of the would-be weatherwise who understood it very 

 imperfectly and were thoroughly incompetent to use it. The 

 successful often made the wildest claims for it, and those less 

 skillful or fortunate were active in denunciations, even derid- 

 ing it as "an optical illusion strengthened by long practice." 

 Secondly, there were real difficulties in the way of its applica- 

 tion. Nearly all observations of the rainband have depended 

 upon eye estimations of its intensity, unsatisfactory at their 

 best and doubly so when complicated by long intervals, clouds, 

 and widely varying conditions of illumination. The rainband 

 is a small object in the instruments generally used for the pur- 

 pose, and it is no easy matter to compare it with such a vague 

 and variable thing as a mental scale of blackness. The refer- 

 ence to the Praunhofer lines is open to nearly as much objec- 

 tion, since their apparent intensity is liable to vary with the 

 light, width of slit and general condition of the spectroscope 

 to an extent which renders them very uncertain standards for 

 scientific purposes. 



The first mentioned cause is an unfailing concomitant of 

 every discovery that smacks of popular science ; the second is 

 due only to the nature of the observation itself and can be 

 removed with comparative ease. The object of this paper is 

 to call attention to a device by which something like quanti- 

 tative accuracy can be secured in the % study of the rainband, 

 and to the general methods which appear to give the best re- 

 sults in this investigation, which, it is to be hoped, will prove 

 of permanent service to meteorology. 



The desideratum is evidently a definite and trustworthy 

 scale of variable blackness, extending over quite a wide range, 

 and easily comparable to the rainband. A direct photometric 

 measure of the intensity of the absorption band would of 

 course be valuable, but is an observation too complicated and 

 delicate for everyday use. The measurement could be made 

 however by using a standard of light as one of the sources of 

 illumination in the spectro-photometer of Yierordt (" Die An- 



