FLAMES OF COMPOUNDS OF CARBON AND HYDROGEN. 421 



doubt as to their accuracy ; and since the preceding paper was read, I have made 

 a much more extensive series of experiments, than the limited time I can devote 

 to such researches, had then enabled me to overtake, involving the determination, 

 with more or less accuracy, of the positions of all the bright lines in the carbo- 

 hydrogen spectrum, whose presence I have been able to detect. These experi- 

 ments, with some account of my methods of observation, I have deemed it desirable 

 to append to the preceding paper. 



Methods of Observation. 



The theodolite A, fig. 2, which was used in measuring the deviations of the 

 refracted rays, has a limb 7*5 inches in diameter, with two verniers reading 10", 

 and a telescope B, of 1-6 inch aperture, furnished with a parallel wire micro- 

 meter. The stage carrying the prism P, furnished with screws to render its faces 

 perpendicular to the divided circle, was mounted over the centre of the theodolite ; 

 and, in order to avoid parallax, the object viewed was an extremely narrow slit 

 placed in the principal focus of the object glass of a 30 inch telescope C L, 

 which thus acted as a collimator.* The telescope rested in Ys, in a solid cast- 

 iron stand, D D, which also carried the theodolite : so that the collimator pre- 

 served an invariable position in relation to the theodolite, notwithstanding any 

 instability of the floor of the room in which the observations were made ; and the 

 zero of the circle was found to remain exceedingly constant. A diaphragm with 

 a vertical slit was placed before the collimator lens, so as to limit the aperture in 

 the plane of refraction to 0*4 inch, and thus to allow only a nearly central pencil 

 of rays to fall on the prism. Any errors, which might have arisen, either from 

 imperfect adjustment of the collimator to sidereal focus, or from defective aplana- 

 tism in its lens, were thus avoided as much as possible. 



The deviations of the refracted rays were observed first to the right, and then 

 to the left, — the prism being always adjusted to its position of minimum devia- 

 tion, — so that the difference of the readings of the verniers in the two positions of 

 the prism, gave double the minimum deviation of the refracted rays. 



The angle of the prism was ascertained, by first turning it with its edge to- 

 wards the object glass of the telescope, as represented in fig. 3, where ABC is the 

 prism, and T the telescope. The stage carrying the prism was rigidly connected 

 with the telescope, so that when the telescope was moved, the prism moved along 

 with it ; and being left undisturbed, the inclinations of its faces AB, AC, to the 

 line of collimation of the telescope remained invariable. The telescope was then 

 turned, until the image of the illuminated slit of the collimator, seen by reflection, 

 successively in the two faces of the prism, was made to coincide with the tele- 

 scope wires ; and, at each intersection, the verniers were read off. The difference 



* I have described this mode of observation in my paper on the Ordinary Refraction of Iceland 

 Spar, Edin. Trans., vol. xvi. 



