•118 Dr. J. H. Gladstone on the Violet Flame of many Chlorides. 



and one iodide, and shows the bearing of his observations on 

 solar chemistry and other questions. 



Now his experiments, though very valuable and suggestive, 

 are far from giving a complete story \ and they recalled to my 

 recollection some experiments of my own, the record of which 

 had never proceeded further than my note-book, but which 

 evidently were elucidated by, and were capable of elucidating 

 those of the German physicist. I have repeated and extended 

 these during the last few days, and I send you the results, hoping 

 to lead other experimenters into the same field, and perhaps to 

 return to it myself when I have greater leisure. 



The fact is that the majority of the lines represented in Alex- 

 ander Mitscherlich's diagram of the flame of chloride of copper, 

 and which, byihe way, are represented in Prof. W. A. Miller's dia- 

 gram of the same in the Philosophical Magazine for August 1845, 

 are common to a large number of chlorides if they are sufficiently 

 heated. In the whole range of spectrum-analysis there is no 

 fact better known than that common salt gives rise to the yellow 

 double line D ; yet it is popularly known that if salt be thrown 

 on to the red-hot coals of an ordinary fire it produces violet 

 flames. This is beautifully seen when old ship timber is burnt. 

 These flames, when analysed, are found to consist, not of the 

 yellow light, but of three groups of lines ; the first green, and 

 extending to the fixed line b ; the second bluish green, and blue, 

 lying on either side of F ; and the third violet, stretching from 

 midway between F and G to a little bej^ond G. When examined 

 by a spectroscope of no great power, these three groups have a 

 close resemblance to one another, each consisting of four lines 

 about equidistant, of which the two middle ones are brighter 

 than the outer ones ; but when they are more carefully examined 

 with a narrow slit, the lines of the second and third groups at 

 least appear as bands of a certain width, and are even resolvable 

 into double bands, of which the more refrangible are the nar- 

 rower and fainter. That -they are identical with the lines of 

 chloride of copper was proved, not only by the identity of their 

 appearance, but by angular measurement of the more prominent 

 ones. No more refrangible rays are seen when this light is 

 examined by a spectroscope the lenses and prism of which are 

 made of quartz. 



Many other chlorides also give the same light when sufficiently 

 heated. The flame of a spirit-lamp is enough to produce it with 

 the chloride of copper; and the chloride of platinum or of gold, 

 if put into such a flame, gives a bright momentary flash of the 

 violet light, which, when analysed by the prism/ is resolved into 

 the same bands, those from the platinum-salt exactly coinciding 

 with those from the copper salt when the two are seen side 



