78 Intelligence and Miscellaneous Articles. 



metal. If, as supposed, this sound is characteristic of the crystal- 

 line structure of metals, it may afford a means of great practical 

 use, whereby, by the sound a metal emits, we may draw conclusions 

 as to its texture, and hence its fitness for certain purposes ; or by 

 the sound emitted by a beam when bent we may draw conclusions 

 as to its safety, the microphone or other appliance being called in 

 to aid us where the sounds are exceedingly weak. — Proceedings, 

 Asiatic Society of Bengal, February 1881, communicated by the Author. 



DISCUSSION OF THE THEORY OF THE THREE FUNDAMENTAL 

 COLOUR-SENSATIONS. BY A. ROSENSTIEHL. 



The notion of the three fundamental colour-sensations arose 

 from the study of the properties of the eye which is imperfectly 

 organized in regard to the perception of colours. The theory which 

 connects the observed phenomena leads to an hypothesis on the 

 structure of the normal eye. 



But since Maxwell's experiments* on the solar spectrum, more 

 than twenty years since, this subject has not again been the object 

 of any investigation, the experimental method for studying the 

 laws of colour- vision upon the normal eye being wanting. By 

 determining, with the aid of rotating disks, the distribution of the 

 complementary colours in a chromatic circle, I believe I have sup- 

 plied that deficiency. The position of the three colours correspond- 

 ing to Young's fundamental sensations has been thereby deter- 

 mined with much more precision than was hitherto possible. 

 Young's theory receiving from this fact a support which takes it 

 out of the domain of hypothesis, it has appeared to me useful to 

 discuss its consequences, and to point out those which are sus- 

 ceptible of exact experimental verification. It will follow from this 

 discussion, on the one hand, that certain properties attributed to the 

 primary colours do not belong to them exclusively, and, on the other, 

 that their true distinctive character has not yet been enunciated. 



The primary colours (that is, those corresponding to the funda- 

 mental sensations) possess, by their very definition, the following 

 properties : — 



1. On being mixed two by two, they produce all the colours percep- 

 tible to our eye. 



This property belongs to all the colours which are not comple- 

 mentary; but at the same time the sensation of white will be 

 produced ; so that the following limiting condition must be added : — 



Tliey produce at the same time the sensation of white in a less degree 

 than the other colours. 



I have already, indirectly, made use of this very important 

 character t. Its experimental verification, however, presents great 

 practical difficulties, because the quantity of white light emitted by 

 a coloured surface cannot be measured with precision. It would 

 be necessary to execute a chromatic circle with all its colours of 



* Proc. Royal. Soc. vol. x. pp. 404-409 (1860). 



t Comptes Rendus, t. xcii. p. 357 ; Phil. Mag. April 1881, p. 305. 



