TRANSACTIONS OF SECTION D. 731 



seems to me the only intelligible theory of smell yet offered. But it must be 

 admitted that a theory of smell such as that advanced by Ramsay involves a more 

 subtle conception of the molecular vibrations in nerve fibrils than is required in 

 the case of hearing. It involves the conception that musk, camphor, and similar 

 substances produce their characteristic qualities of smell by setting up nerve 

 vibrations probably of dififerent frequencies and different complexities, We shall 

 see what bearing this may have on the theory of colour sense, to which I now pass. 



No impressions derived from external Nature yield so much joy to the 

 mind as our sensations of colour. Pure tones and perfect harmonies produce 

 delightful sensations, but they are outrivalled by the colour effects of a glorious 

 sunset. Without our sense of colour all Nature would appear dressed in bold 

 black and white, or indifferent grey. We would recognise, as now, the beauty of 

 shapely forms, but they would be as the cold engraving contrasted with the 

 brilliant canvas of Titian. >The beautiful tints we so readily associate with natural 

 objects are all of them sensations produced in our brain. Paradox though it 

 appear, all Nature is really in darkness. The radiant energy that streams from a 

 sun is but a subtle wave-motion, which produces the common effects of heat on all 

 bodies, dead or living. It does not dispel the darkness of Nature until it falls on 

 a living eye, and produces the sense of light. Objective light is only a wave 

 motion in an etherial medium ; subjective light is a sensation produced by mole- 

 cular vibration in our nerve apparatus. 



The sensorj' mechanism concerned in sight consists of the retina, the optic 

 nerve, and the centre for visual sensation in the occipital lobe of the brain. In the 

 vertebrate eye the fibres of the optic nerve spread out in the inner part of the 

 retina, and are connected with several layers of ganglionic cells placed external to 

 them. The light has to stream through tlie fibres and ganglionic layers to reach 

 the visual cells — that is, the nerve terminals placed in the outer part of the retina. 

 They may be regarded as epithelial cells, whose peripheral ends are developed into 

 peculiar rod and cone-shaped bodies, while their central ends are in physiological 

 continuit}' with nerve fibrils. Each rod and cone consists of an inner and an outer 

 segment. Tlie outer segment is a pile of exceedingly thin, transparent, doubly 

 refractive discs, colourless in the cone, but coloured pink or purple in the rod. In 

 man, the inner segment of both rod and cone is colourless and transparent. Its 

 outer part appears to be a compact mass of fine fibrils that pass imperceptibly into 

 the homogeneous-looking protoplasm in the shaft of the cell. Owing to the 

 position of the rods and cones, the light first traverses their inner, then their outer 

 segments, and its unabsorbed portion passes on to the adjacent layer of dark- 

 brown pigment cells by which it is absorbed. It is not necessary for me to discuss 

 the possible difi'erence of function between the rods and cones. I may simply say 

 that in the central part of the yellow spot of the retina, where vision is most 

 acute, and from which we derive most of our impressions of form and colour, the 

 only sensory terminals are the cones. • A single cone can enable us to obtain a 

 distinct visual impression. If two small pencils of light fall oh the same cone the 

 resulting sensory impression is single. To produce a double impression the 

 luminous pencils must fall on at least two cones. That shows how distinct must 

 be the path pursued by the nerve impulse from a visual cell in the yellow spot to a 

 sensory cell in the brain. The impulses must pursue discrete paths through the 

 apparent labyrinth of nerve fibrils and ganglion cells in the retina to the fibres of 

 the optic nerve. I now pass to tlie physical agent that- stimulates the retina. 



When a beam of white light is dispersed by a prism or diffraction grating, the 

 ether-waves are spread out in the order of their frequency of undulation. The 

 undulations of radiant energy extend through a range of many octaves, as Hertz 

 has recently shown, but those able to stimidate the retina ard comprised within a 

 range of rather less than one octave, extending from a frequency of about 395 

 billions per second at the extreme red to about 757 billions at the extreme violet 

 end of the visible spectrum. The ultra-violet waves in the spectrum of sunlight 

 extend through rather more than half an octave. Although mainly revealed by 

 their chemical effects, they are not altogether invisible: their colour is bluish-grey. 

 Ilxe only ojitical — that is, strictly ^(/(ys/cf// — difference between the several ether- 



