50 ALEXANBEE JOHNSON 



In Parkinson's " Optics," second edition, 1866 (a Cambridge book), the same error is 

 contained, not as a direct statement, bnt by implication, for, after describing Newton's 

 experiment with a small aperture, it says, p. 149 : " Instead of a very small aperture 

 Wollaston and Fraunhofer admitted the sun's light through a very narrow slit, the 

 effect of the slit being to give an assemblage of innumerable linear spectra placed side 

 by side." 



Proctor ("Spectroscope," 187t, p. 16) does not seem to be aware that Newton had 

 used a narrow slit, for although he refers to his using an " oblong " and a " triangular " 

 aperture as well as other shapes, yet it appears, from his contrasting these with Wollas- 

 ton's use of a slit as well as from his diagram, that he considered the triangles (equilateral) 

 and the " oblongs " to be about the same size as the round hole also employed by Newton. 

 It appears, however, more definitely from his work on " The Sun" (p. 101, 18Y2) that ha 

 shared the common error. He says : " Wollaston foiind that when, instead of a circular, 

 triangular or oblong aperture, a very narrow slit is employed, light of certain degrees of 

 refrangibility is absent from the solar beam ;" and on the same page he remarks : " This 

 mode of viewing the spectrum bears the same relation to Newton's plan," etc He does 

 not appear to have consulted WoUaston's original paper, for he says : '^ The spectrum seen 

 by "Wollaston was not continuous, but crossed by two dark lines parallel to the slit," 

 whereas Wollaston states that he saw six lines. Curiously enough, Parkinson also says : 

 " Two of the fixed lines, probably E and F, had been discovered by Wollaston previous to 

 the experiments of Fraunhofer." Yet Sir David Brewster (" Optics," 1853, p. 91) says of 

 them : " These six lines are found to correspond with those marked B, D, b, F, Gr and H " 

 [by Fraunhofer]. 



Heath's " Greometrical Optics " (Cambridge, 188*7) alludes (p. 195) to Newton's experi- 

 ments with a small circular hole only, remarking (p. 196) that " the colours will not be 

 thoroughly separated ; the spectrum is then said to be impure." How a pure spectrum 

 may be obtained is described immediately afterwards, without any reference to Newton. 

 I tried to draw attention to this general error by a letter which appeared in ' Nature' 

 in October, 1882, and should hardly have referred to it again had it not been for the recur- 

 rence of the same statement in Sir William Thomson's " Popular Lectures" (vol. i, p. 324, 

 1889), where he says : " Newton never used a narrow beam of light, and so could not 

 have had a homogeneous spectrum " The lecture was on " The Wave Theory of Light," 

 and given in Philadelphia in 1884. 



The weight of Sir William Thomson's name is so deservedly great that this statement 

 by him is likely to greatly extend the prevalence of the error. The republication of the 

 original work, now so difficult to procure for consultation, seems the best way of obviating 

 this and other mistakes concerning it. Meanwhile I make the following extracts from 

 the first edition (1704), in which it will be noticed that Newton used the lens also, 

 although not to make the rays parallel. 



In Prop. 4, Bk. I, of the " Opticks," 1*704, Newton proposes the problem to find a pure 

 spectrum, or, as he words it, " To separate from one another the Heterogeneous Eays of 

 Compound Light." 



After showing at some length (p. 47) why he uses a lens to " diminish the mixture 

 of the Rays," he describes experiment 11, first with a round hole, and afterwards with a slit, 

 as follows : 



