672 TRANSACTIONS OF SECTION E. 
colours. The race which we have used is very sporting, a single plant bearing 
in one inflorescence deep magenta, pale magenta, white with limited rose flush, 
and all but pure white flowers. 
li a petal of each of these flowers be treated with the benzidine reagent, 
it is found that the extent and amount of the oxydase reaction, as measured by 
the distribution and depth of brown coloration indicative of oxydase, coincide 
precisely with the extent and amount of pigmentation. The full-coloured petal 
gives a uniform deep brown reaction, the light magenta a uniform but paler 
reaction, the petal with a limited rosy flush gives a slight reaction, limited to 
the pigmented area, and the all-but-white petal gives none but the slightest 
reaction, and that only in the part of the petal which contained traces of pigment. 
Thus—unless the results are due to a partial inhibition which has eluded our 
attempts at demonstration—it would seem established that the ever-sporting 
habit is due to differences in the amount of oxydase in the diversely coloured 
flowers. 
The Sweet William is also noteworthy in that it contains white races, some 
of which give an oxydase reaction in their petals and some of which give no 
oxydase reaction. Breeding experiments now in progress will decide whether 
or no these white races, like those of Sweet Peas investigated by Bateson 
and Punnett, mated together yield coloured progeny. If so the factors for 
colour, long wandering yet not lost, which meet again in reversionary coloured 
crossbreeds, may prove to be a chromogen factor and an oxydase factor. 
Finally a brief reference must be made to our observations on the periodic 
fluctuation of oxydase in plants. Various observers have noticed that plant 
tissues give the peroxydase reaction much more generally than the oxydase 
reaction. The observations now to be described indicate that this is due to 
the greater stability of peroxydase as compared with the organic peroxide. 
Under certain circumstances a tissue which gives only the peroxydase 
reaction may exhibit the direct oxydase reaction. Moreover, the extent of the 
peroxydase reaction, as judged by the depth of coloration of the reagent, 
varies in similar plants at different times. 
Enquiry into the meaning of these fluctuations led us to the discovery that 
the nature and amount of oxydase contained in a plant tissue varies in an 
orderly manner according to external conditions. 
Among the conditions which determine this fluctuation are light and darkness. 
Plants subjected to normal illumination possess less oxydase than those which 
are kept in darkness. After one or two days’ exposure to darkness plants of 
P. sinensis contain more peroxydase than sister plants kept under normal con- 
_ditions of illumination. Moreover, after such an exposure to darkness tissues 
which under normal conditions give only peroxydase-reactions yield distinct 
oxydase-reactions. 
Whether these phenomena are general among plants we are not yet in a 
position to say; but repeated experiment enables us to vouch for them in the 
case of P. sinensis. Should the results of similar investigations with other 
plants show that this diurnal variation of the oxydase-content of plant tissues 
is of general occurrence, we may perhaps discover therein the means whereby 
many of the phenomena of periodicity exhibited by plants are maintained 
and regulated. We know that the light and darkness of the day and night 
set up rhythms in the plant; for example, that the leaves of various plants 
assume nocturnal and diurnal positions. We know further that the rhythm 
thus established may be maintained for a certain time under uniform conditions 
of illumination. This is the case with the Sensitive plant and many another. 
Animals also exhibit a like periodicity. Thus some years ago Dr. Gamble and 
I showed that certain shrimp-like animals, Hippolyte varians, roll up their 
hrilliant chromatophores at night and assume a sky-blue colour. When daylight 
comes they put on their daytime dress by spreading out the pigment of their 
chromatophores in far-reaching superficial networks. Kept in the dark these 
animals retain for many days this periodic habit, and when the hour of night 
arrives, although they have no light to tell it by, they lay aside their daily 
garb and put on the uniform of night. So also the plant-animal Convoluta 
roscoffensis, which lives on the seashore, orders its behaviour by the sun and 
moon, It lies on the sand till the waves of the making tide are upon it, and 
