274 M. R. Ruhlmann on the Alteration produced by Heat 



has been exactly determined, then the refracted image of the 

 slit illuminated by the sodium-name is brought to coincide with 

 the vertical thread of the telescope. It is well known that at 

 the minimum refraction the rays, on emerging, diverge in the 

 same manner as if they had proceeded directly from the slit j so 

 that, since the difference in the paths is very small, the slit may 

 be seen directly and distinctly. If the instrument is accurately 

 adjusted, the refracted and direct images of the slit must appear 

 at the same height in the telescope. If the difference is consi- 

 derable, the interval must be measured by means of the vertical 

 circle, and the angular reading divided by the cosine of the dif- 

 ference in height. 



The essential difference between the method adopted by me 

 and that employed by Fraunhofer, Baden Powell, and others 

 consists in the complete separation of my prism from the tele- 

 scope, and its being placed upon a graduated circle so that its ro- 

 tation can be measured. I chose this arrangement partly be- 

 cause I could thus warm the prism at will, and keep it warmed 

 without altering its position, and finally because by this means 

 I could introduce continual determinations of the refracting 

 angle by means of reflexion-observations. Moreover this me- 

 thod is rather to be preferred to the others for measuring the 

 alteration of the index of refraction with the temperature. 



On altering the refractive power of the medium with which 

 the hollow prism is filled, the angle of least refraction is changed. 

 But since the telescope receives and collects parallel rays as long- 

 as any part of the light which leaves the prism falls within the 

 cone which can be drawn from the eyepiece to the object-glass, 

 it is clear that the refracted image of the slit can be included, 

 even under great alterations of the angle of least refraction. 



The observations divide themselves under two principal heads — 

 the determination of the refractive angle of the prism, and the 

 determination of the angle of least deviation. 



(1) The measurement of the angle of the prism. — A prism being 

 fastened to a graduated moveable circle (arranged as described), 

 and an auxiliary telescope at the same level being directed to it, 

 distant objects can be seen reflected on its surface. The angular 

 determination is found simply from the difference of the read- 

 ings on the graduated circle when one and the same distant 

 object appears reflected on one or the other surface of the prism. 



(2) The determination of the angle of least deviation. — By the 

 arrangement adopted, the angle of least deviation is not directly 

 measured upon the universal circle, but a correction must be 

 introduced in the reading in consequence of the excentric posi- 

 tion of the prism. If (fig. 3) S denotes the slit, P the prism, 

 the centre of the universal circle, and 3 the angle of refrac- 



