1908] CROCKER AND KNIGHT—CARNATIONS - 269 
to determine the toxic limits of methane, carbon monoxid, acetylene, 
hydrogen, carbon bisulfid, and benzene to the buds and flowers. 
As would be expected, hydrogen was perfectly neutral when it com- 
pletely displaced the nitrogen of the air. In all the other constit- 
uents here mentioned, the toxicity was such that in the least amount 
of illuminating gas necessary to kill the bud no one is concentrated 
enough to reach sy Of its toxic limit. It is very probable, therefore, 
that these constituents play no part in determining the toxic limit 
of illuminating gas. It has already been stated that the absorption 
of hydrogen sulfid and sulfur dioxid does not modify the toxicity of the 
gas. This leaves, then, ethylene, the higher homologues of ethylene 
and acetylene, and certain aromatic sulfur compounds to account for 
the toxicity of the gas. All these substances except ethylene exist 
in very small percentages in illuminating gas. All evidence in the 
following experiments also points to the conclusion that there is 
fnough ethylene in the gas to account for its toxicity. 
The small greenhouse in which entire potted plants were exposed 
'o the action of gas had a capacity of 1.69%™. In order to make 
comparisons easy between buds of the same size on the plants exposed 
and on the checks, corresponding buds were tagged with the same 
numbers. We need describe only one of these experiments. Potted 
Plants of the Boston Market were put into the small greenhouse in 
the evening and 2 liters of gas were run in at the end opposite the 
Plants, allowing a gradual distribution by diffusion. The plants 
"ere taken out the next morning to prevent injury by high tempera- 
ture. The following evening the plants were returned to the enclos- 
ure and left for 60 hours (the following two days being cloudy). At 
the time they were put in, 4 liters of gas were run in, and the same 
‘mount Was added 48 hours later, there being at that time no per- 
ceptible smell of gas in the chamber. This experiment served to 
4 (r) that the vegetation is far less sensitive to gas injury than the 
; for there was no apparent injury to the vegetation; (2) plants 
remained vigorous , put out new buds, and later produced other flowers. 
€ oldest buds (those showing color and just ready to open) and the 
Jos 8est buds (those less than o.6°™ in diameter) were the ones most 
_ Many of the medium-sized buds, however, escaped death, 
although retarded considerably in their growth. The older buds 
