25 
in order to obtain complete darkness, and to diminish the influence 
of the variations of temperature when light is not required. By 
sinking it in the ground, by the thickness of its walls, and by the 
covering of its exterior surfaces with straw, mats, etc., the same 
fixed degree of temperature could be obtained as in a cellar. The 
vaulted building should have an underground communication with 
a chamber containing the heating and the electrical apparatus. The 
entrance into the experimental hothouse should be through a passage 
closed by a series of successive doors. The temperature should be 
regulated by metallic conductors, heated or cooled at a distance. 
Engineers have already devised means by which the temperature of 
a room, acting on a valve, regulates the entry or exit of a cértain 
amount of air, so that the heat regulates itself. Use could be made of 
such an apparatus when necessary. 
Obviously, with a hothouse thus constructed, the growth of plants 
could be followed from their germination to the ripening of their 
seeds, under the influence of a temperature and an amount of hght 
perfectly definite in intensity. It could then be ascertained how heat 
acts during the successive phases from sowing to germination, from 
germination to flowering, and from this on to the ripening of the 
seed. For different species various curves could be constructed to 
express the action of heat on each function, and of which there are 
already some in illustration of the most simple phenomena, such as 
germination, the growth of stems, and the course of the sap in the 
interior of certain cells. We should then be able to fix a great num- 
ber of those minima and maxima of temperature which hmit phys- 
iological phenomena. Indeed, a question more complicated might 
be investigated, toward the solution of which science has already 
made some advances, namely, that of the action of variable tempera- 
tures; and it might be determined if, as appears to be the case, these 
temperatures are sometimes beneficial, at other times injurious, ac- 
cording to the species, the function investigated, and the range of 
temperature. The action of hight on vegetation has given rise to 
the most ingenious experiments. Unfortunately these experiments 
have sometimes ended in contradictory and uncertain results. The 
best ascertained facts are the importance of sunlight for green col- 
oring, the decomposition of carbonic-acid gas by the foliage, and 
certain phenomena relating to the direction or position of stems and 
leaves. There remains much yet to learn upon the effect of diffused 
light, the combination of time and light, and the relative importance 
of light and heat. Does a prolonged light of several days or weeks, 
such as occurs in the polar regions, produce in exhalation of oxygen, 
and in the fixing of green matter, as much effect as the light distrib- 
uted during twelve-hour periods, as at the equator? No one knows. 
In this case, as for temperature, curves should be constructed, showing 
the increasing or diminishing action of light on the performance of 
each function; and as the electric light resembles that of the sun, 
we could in our experimental hothouse submit vegetation to a con- 
tinued hght. 
A building such as I propose would allow of light. being passed 
through colored glasses or colored solutions, and so prove the effect 
of the different visible or invisible rays which enter into the compo- 
sition of sunlight. For the sake of exactness nothing is superior to 
the decomposition of the luminous rays by a prism, and the fixing the 
