152 CO-OPERATION AND THE ORIGIN OF FLOWERS. 
sunshine, 8873 distinct and separate flowers have been picked 
in a square of stony soil measuring 30 centimetres each 
way.J2 
Near Alexandria, in Egypt, on certain dry limestone 
hillocks, I once saw a profusion of rich reds, glowing crim- 
son, golden yellow, blues and purples of quite indescribable 
beauty; the plants which produced them were all small and 
quite insignificant. Other places also remarkable for 
similar gorgeousness were the sandhills near Concepcion in 
Chili, the Karroo in South Africa, and Switzerland. 
All these habitats correspond to that which we have 
assumed as the primeeval flower climate. 
In Scotland, it is my impression that on the seacoast 
on dry, wind-exposed rocks, the maximum of richness and 
profusion in colour will be found.* 
Moreover, though this is not the place for any detailed 
and complete account of anthocyan, it is (I think) safe to say 
that under strong sunshine, which might be injurious to the 
delicate contents of living cells, certain organised bodies in 
the protoplasm produce intense respiration. Under this ex- 
cessive respiration the anthocyans (which are glucosides) are 
formed from the sugars already existing in the protoplasm. 
It is also usually believed that the red anthocyan so produced 
acts as a light screen, thus protecting the underlying cell-con- 
ce 
tents: In this case, then, the “ evil’ directly produces a 
remedy against itself. 
It is well known that respiration is more intense in 
flowers, and that their temperature is often considerably 
higher than that of the air surrounding them. 
In certain interesting experiments of Colonel Rawson, 
it was found that the flowers of Tropeolum majus became 
purple when exposed only to overhead light, whilst they were 
red under light at medium elevations, and yellow if the sun 
was only allowed to reach them when close to the horizon.4 
These observations should be repeated and extended.t The 
* Here also detailed observations are required (fourth series). 
t More experiments are required (fifth series); they would pro- 
bably be best carried out in a small garden. 
