264 BOTANICAL GAZETTE [OCTOBER 
effect of illuminating gas upon the flowering carnation. These 
growers claimed to have had heavy losses from gas that seeped from 
defective pipes through the ground into the greenhouses. In some 
cases it is claimed that the losses occurred during cold weather, when 
little ventilation was possible and when the ground was frozen, so 
that upward diffusion from the defective pipes was hindered and 
thereby lateral diffusion fostered. In all these cases it is claimed that 
the injuries ceased with the repair or removal of the pipes. 
Upon looking up the literature it was found that no accurate 
determinations were made upon the effects of illuminating gas and 
its constituents upon flowers, and that in no case have the toxic limits 
and relative toxicity of the several main constituents been deter- 
mined. In short, it is not known in any case whether the toxic limit 
of the gas is determined by the action of one constituent or by the 
combined action of several. To answer these questions is the put 
pose of the investigation here reported. 
This paper will deal entirely with the buds and flowers of the 
carnation, describing in detail the effects and toxic limits of illuminat- 
ing gas and ethylene. A later paper will give in detail similar data 
for the other main constituents of illuminating gas, as well as escribe 
the effects of illuminating gas and all its main constituents upo? the 
vegetation of the carnation. The work naturally falls into these 
two divisions, for, as will be shown by experiments described later, 
the flowers are far more sensitive to illuminating gas than is the 
vegetation, and the toxic limit of the gas on the flowers seems (from 
all the evidence of our experiments) to be entirely determined by the 
ethylene it contains. 
To determine the relative sensitiveness of buds and flowers oP sa 
one hand, and the vegetation onthe other, as well as the relat 
sensitiveness of buds and flowers of different ages, one series © 
experiments was carried on by exposing entire potted plants 1 = 
atinosphere containing small proportions of gas. This was done by 
setting the plants into an air-tight greenhouse within the laboratory 
greenhouse, and then running desired quantities of gas into ee 
tight greenhouse. This sort of experiment has some serious faults. 
It does not determine whether the flower is affected directly caer! 
gas contained in the air about it, or whether the effect is in . 
