530 SECTIONAL TRANSACTIONS.—H, I. 
main aspects are dealt with in detail, and are illustrated on gramophone 
records of primitive music. 
(1) The sociological aspect. Music as general enjoyment and as social 
privilege in primitive society. Professional musicians. Music as a central 
factor in religious ceremony. 
(2) The psychological aspect. General laws established by the psychology 
of sound can explain certain widespread characteristics in primitive music. 
Thus the fluctuation of intervals can be derived from psycho-physiological 
factors, or the development of a certain primitive polyphony (the so-called 
‘ Parallel-Organum ’) appears to be due to the elementary psycho-physio- 
logical phenomenon of the ‘ similarity of tones.’ 
(3) The aspect of historical connection. Historical relations, borrowing 
and diffusion, have to account, in certain cases, for corresponding musical 
styles in different areas, thus checking very often the psychological aspect 
(e.g. the ‘ historical’ explanation of the Parallel-Organum suggested for 
certain areas). Musical instruments become here specially important. 
Built on principles of physics, and bound up with measures, they offer 
most accurate and objective criteria for establishing cultural relations 
(e.g. the ethnology of the African ‘ marimba ’). 
(4) The racial (biological) aspect appears to be tangible through music—i.e. 
through certain musical characteristics which are deeply rooted in biological 
factors (e.g. the motoric type of man). It is tried to illustrate this as yet 
very tentative approach on the melodic motion-type characteristic for the 
songs of two racial groups: the American Indian and the Austronesian. 
SECTION I.—PHYSIOLOGY. 
Thursday, September 7. 
Dr. F. W. Epripce-Green, C.B.E.—A criticism of Roaf’s theory of 
colour vision (10.0). 
Whilst Roaf’s work on colour vision and very fair criticisms of the work 
of others deserve great praise, his theory presents the same difficulties as 
any form of the trichromatic theory. ‘There is no evidence of any trichro- 
matic light perceiving apparatus in the human retina. Houstoun has shown 
that the trichromatic theory is mathematically untenable. The chief objec- 
tion to it, however, is that like every other theory but mine, it will not explain 
colour-blindness, particularly those facts predicted by my theory. How, for 
instance, can the fact that 50 per cent. of the dangerously colour-blind pass 
the wool test be explained, or that 90 per cent. of the dangerously colour- 
blind agree with the normal white equation, though they may make an 
anomalous white equation ? 
A dichromic or trichromic makes mistakes through defective discrimina- 
tion and not through the defect of a light perceiving substance when there 
is no shortening of the spectrum or defect in light perception. A man with 
shortening of the red end of the spectrum may recognise a very feeble red 
of shorter wave-length, so this condition cannot be due to a defect in a light 
perceiving substance which is affected by all rays of the spectrum. 
Joint Discuss1on with Section J (Psychology, g.v.) on Disorientation 
and vertigo (10.30). 
