214 



p. 119.) M. Libes {llistoire Philosophique des Progres dela Phy- 

 sique, Paris, 1810-1812), mentions Papin and Hauksbee as the only 

 claimants of the double pump (t. iii., p. 56j ; and adds, that Cotes, 

 the mathematician of Cambridge, who was contemporary with Hauks- 

 bee, i-egarded Papin as the author of the invention. If there are no 

 claimants known even to Continental historians of science, but 

 Hauksbee and Papin, the latter, whose instrument was constructed 

 more than twenty years before Hauksbee's, is entitled to the whole 

 honour due to the inventor of the double air-pump. 



2. On the Classification of Colours. Part II. By Professor 

 J. D. Forbes. (See p. 190.) 



The object of this paper is chiefly one of nomenclature. Every 

 one has felt the difficulty of describing with the precision the innu- 

 merable hues which occur in nature and in art ; and which it is equally 

 desirable for the optical philosopher, the artist, and the manufacturer, 

 to be able to refer to in a clear and definite manner. But such a 

 nomenclature or classification must proceed upon some admission as 

 to the manner of compounding complex hues out of simple ones ; and, 

 therefore, the author first treats of the (so-called) Primary Colours. 

 He admits it as highly probable, that all known colours may be 

 formed out of Red, Yellow, and Blue ; although, when we attempt 

 to compound pigments, we have a very notable loss of light, and also 

 an unavoidable impurity, which is most visible in the compound tints. 

 The author, in passing, endeavom-s to explain clearly why the union 

 of pigments never can produce a perfect white, although the coloured 

 light of the spectrum does so ; for, by adding blue light to yellow 

 light, we not only change the colour, but we increase the illumina- 

 tion ; whereas, by adding a blue to a yellow pigment, whilst we change 

 the colour, we at the same time reduce the luminousness of the sur- 

 face, the blue particles being far less reflective than the yellow ones. 

 Inferring from Newton's empirical rule, the quantities of red, yellow, 

 and blue light, which should combine to make white light ; and 

 adopting Lambert's results as to the reflective powers of the brightest 

 picrments, the author concludes, that the mean illumination of a disk 

 put in rapid revolution, and containing coloured sectors, will be 4-57 

 times less than if it reflected the whole incident light, or it will re- 

 flect only about half the light which white paper does under the same 



