282 HISTORY OF SCIENCE. 



the yellow ray which forms the brightest part of the spectrum, we may 

 in any given case determine by what angle its direction, as it emerges 

 from the prism, differs from the direction of the original ray. This 

 angle may represent the mean refraction of the ray, and may be called 

 the angle of deviation, or simply the "deviation" Again, the angle 

 between the directions of the emergent extreme red and violet rays, 

 which represents the spreading out of the colours, is termed the angle 

 of dispersion, and this we may call, for shortness, the " dispersion" 

 Now, the conclusion which Dollond and others had drawn from New- 

 ton's experiments was that the dispersion and deviation have one and 

 the same fixed proportion to each other in every case. 



The reasoning of Klingenstierna, the Swedish geometer already 

 alluded to, set Dollond on the true experimental track, and in 1757 

 he "began a series of experiments which he prosecuted, as he says, " with 

 a resolute perseverance," devoting nearly a year and a half to this one 

 object, until his investigations led him to a most interesting and im- 

 portant discovery. One of his first experiments was to form a prism 

 with water contained in a vessel with glass sides, so arranged that the 

 angle between these could be adjusted to any desired degree. In 

 this vessel he placed a glass prism, with its refracting angle uppermost. 

 Thus he had virtually two prisms, one of glass and ene of water, with 

 their angles in opposite directions, so that their united action tended 

 to destroy each other's effects ; and by trial he so adjusted the inclined 

 faces of the water prism, that the refractive effect of the combination 

 was completely destroyed ; that is, objects viewed through the prism 

 were seen in their ordinary positions. In this case, then, the " devia- 

 tion" was nil; and therefore, if Newton's principle were true, the 

 " dispersion " should have been also nil. So far was this from being the 

 fact, that the objects were seen as much surrounded by coloured fringes 

 as if they had been viewed through the glass prism alone. Dollond's 

 next step was to place in the water prism a glass prism of a small 

 refracting angle, and then he changed the angle of the water prism 

 until he found one at which the objects were seen free from colour, and 

 yet there remained a considerable amount of refraction. Having thus 

 proved by experiment the possibility of constructing achromatic lenses, 

 Dollond proceeded to try experiments with prisms prepared of various 

 substances, and was fortunate enough to find in two different kinds of 

 glass precisely what he required. The common glass, which goes by 

 the name of crown-glass, has a refractive power which is but slightly 

 inferior to that possessed by the denser kind of glass called flint-glass, 

 while the dispersive power of crown-glass is greatly inferior to the 

 dispersive power of flint-glass. The refractive powers of these kinds of 

 glass are as 20 : 21, while their dispersive powers* are as 14 : 2 r. It is 

 therefore possible to make two prisms, one of flint-glass and the other 



* The measure of the dispersive power is the ratio of the angle of dispersion to the 

 angle of deviation. 



