452 Proceedings of Royal Society of Edinhurgh. [jult 16, 
limits of time and intensity, any set of fibres is entirely unstimu- 
lated while the others are excited. We must, therefore, assume 
that, although the beginning of each curve may be a straight line, 
the rest of the curve is not straight, or at any rate not in the same 
straight line as the initial part. 
Fig. 4 represents the course of the curves which most naturally 
explains our phenomena. 
Here the proportion 
AiEj: A 2 B 2 : A 0 B 3 
( 1 ) 
will in general be different from 
• ^ 2^2 * ^8^3 * * 
( 2 ) 
and this again from 
* ^2 ^2 * ^3 ^3 j 
(3) 
which also differs from 
(4) 
We may take (1) as the proportion which gives the impression of 
white light. Eed, green, and violet have all reached their maximum 
here. Since red reaches its maximum first, its curve is drawn as a 
straight line parallel to the abcissa for some distance to the left of 
ordinate A^Bj. 
Proportion (2) wonld give a preponderance of violet, (3) of green, 
(4) of red. These curves represent the course of the excitation for 
a single flash. For a series of flashes following one another at an 
interval not much greater than that necessray for fusion, the ordinate 
corresponding to the length of each flash will still be the mean 
ordinate of the compound curve. 
In fig. 5, the continuous line represents the course of the excita- 
