} 
j 
" 
a 

oo . 
June 16, 1923] 
799 

are a low-pressure plant for supplying air to keep back 
the water ; a high-pressure plant to supply compressed 
water supply for grouting and washing ; an electric light 
and power supply ; and transportation plant. A useful 
chapter is one on the working force needed, the rate of 
progress in different cases, and the cost. 
The authors give a theory of the stresses in the 
tunnel lining, a subject hitherto far too much neglected, 
designs having followed rule-of-thumb methods. This 
is not a suitable occasion for discussing a mathematical 
theory. The mode of treatment is unusual, but the 
results are interesting. The authors seem to underrate 
the erection stresses due to the weight. The most 
important external load is the earth pressure. The 
theory of earth pressure of Rankine is adopted, in 
which ¢ depending on the angle of repose is the ratio of 
the “active ”’ horizontal pressure to the vertical pres- 
sure and 1/c the ratio of the “ passive ” resistance of 
the earth, if the structure presses against it and is on 
the point of displacing it. But the statements (p. 76) 
that if the active horizontal pressures are not suffi- | 
ciently large in relation to the vertical pressures the 
tunnel will have a tendency to deflect horizontally, and 
(p. 53) that if % lies between ¢ and 1/c the tunnel lining 
will not be subject to a moment, require more justifi- 
cation. 
An interesting chapter is that on compressed-air 
illness and the precautions to prevent it. The cause is 
the absorption of an excess of nitrogen by the blood— 
disengaged if the pressure is reduced. The cure is 
careful limitation of the period of work and slow decom- 
pression. If in spite of precautions cases of illness 
occur, the remedy is to recompress and decompress more 
slowly. For this a hospital lock is provided. 
W. C. U. 
Colour Vision and Colour Vision Theories. 
(1) Colour Vision: A Discussion of the Leading 
Phenomena and their Physical Laws. By Prof. W. 
Peddie. Pp. xii+208. (London: E. Arnold and 
Co., 1922.) 12s..6d. net. 
(2) Colour and Methods of Colour Reproduction. By 
Dr. L. C. Martin. With Chapters on Colour Printing 
and Colour Photography, by William Gamble. Pp. 
xiii+187. (London, Glasgow and Bombay: Blackie 
and Sons, Ltd., 1923.) 12s. 6d. net. 
HERE are no subjects on which discussion and 
demonstration are more needed than those of 
vision and colour vision. The trichromatic theory as 
presented by Helmholtz was the best theory in relation 
to the facts known at the time, but the difficulties of 
No. 2798, VOL. 111] 

' the theory were thoroughly recognised by him. That 
_it was a theory and not a fact was stated by Helmholtz. 
air for working rock drills and other tools ; a service 
That all colours can be matched by a mixture of three 
| selected simple colours is a fact, but the statement 
that there is an underlying trichromatic basis is not 
only not a fact, but it is also not supported by any fact 
which cannot be explained in another way, and there 
is the most conclusive evidence that this is not the 
case, while another explanation is completely con- 
sistent with the facts. The state of chaos existing in 
many minds with regard to colour vision is due to the 
assumption that the trichromatic theory is a fact. If 
the theory be denied there is no evidence for it, and this 
was known to Helmholtz. 
(x) Prof. Peddie’s book is an uncompromising accept- 
ance of the theory as a fact. The book is an admirable 
exposition of the functions of three variables, but no 
attempt is made to answer any of the objections to 
the theory or to show how the theory is consistent 
with known facts. In this respect the author differs 
from other writers, who admit that something more 
isrequired. Ifweregard colour perception as developed 
secondarily to light perception, as it undoubtedly was, 
we can form a series from total colour blindness to 
super-normal colour perception. The colours differing 
most physically in wave-length being first discriminated, 
these gradually approached each other until green was 
discriminated in the centre as a new colour, then yellow, 
then blue, then orange, and lastly indigo. The ex- 
planation, therefore, why red and green make yellow 
when mixed is, that yellow having replaced the red- 
green of a previous state of development, the colour 
perception is not sufficiently developed to discriminate 
/between a mixture of red and green and simple yellow. 
No two accounts of the trichromatic theory agree, 
‘and the theory is loaded with subsidiary hypotheses, 
many of them quite inconsistent with each other ; that 
«is to say, one will explain one set of facts but not another 
set of facts for which a different arrangement is required. 
Space will not permit of more than a few of the very 
long list of objections to the trichromatic theory being 
given here. If aymixture of spectral lights red, green, 
and violet be made to match a simple white, on the 
_trichromatic theory the internal physiological processes 
should be identical ; but if the eyes be now fatigued 
with a red light containing that used in the mixture, 
about twice as much green will be required in the 
mixed white, the mixture appearing bright green to 
a normal person with unfatigued eyes. Again, if a 
spectrum be viewed with an eye fatigued by looking at 
burning sodium, the yellow will have disappeared from 
the spectrum and the red and green will appear to 
meet, but a feeble red at the end of the spectrum will 
be quite visible. If, however, after looking through a 
