274 BOTANICAL GAZETTE [OCTOBER 
gas. The most delicate application of the blood test (see HEMPEL, 
p- 225) will detect 1 part of carbon monoxid per 40,000. The iodine 
pentoxid test (see HEMPEL, p. 226) is of equal delicacy. If carbon 
monoxid forms 25 per cent. of illuminating gas, these tests will detect 
1 part of illuminating gas in 10,000. . Upon three days’ exposure ! 
part of illuminating gas in 40,000 kills the young buds and the petals 
of the flowers just beginning to open; while 1 part in 80,000 causes 
open flowers to close upon an exposure of twelve hours. 
The so-called “sleep” or closing of the carnation is a source of 
considerable loss to growers and dealers, for flowers that once close 
never again open. This “sleep” is especially likely to occur with cut 
flowers brought into city markets. Some varieties are so disposed to 
react in this way that their cultivation has almost entirely ceased. 
We know several homes lighted with gas where cut carnations can be 
kept only a few hours without “going to sleep.” In one instance 
the displacement of gas lights by electric lights entirely overcame 
this difficulty. Our experiments show clearly that one cause of this 
sleep is traces of illuminating gas (ethylene) in the surrounding atmos- 
phere. : 
STONE (13), WEHMER (8), and others have shown that illuminating 
gas diffuses great distances through the soil, especially if there  ® 
hard-packed or frozen crust over the top. This paper shows the 
extreme sensitiveness of the carnation to this substance. From these 
facts it is evident that carnation growers whose greenhouses are - 
the region of gas pipes must take great precautions against ee 
from this source. It would be interesting to know whether solid 
cement walls set into the ground for some depth on the side — si 
pipes would furnish sufficient protection against leaks of this kind. 
It is clear that, if (as our results seem to indicate) the group of illum 
nants, or more accurately if one constituent of this group (ethylen® 
determines the toxicity of illuminating gas, coal gas is foes 
less toxic than water gas, while oil gas is more toxic than either of 
others; also the toxicity reported by the German investigator | 
used coal gas is less than that shown by the gas of the great Am 
cities. 
While it seems probable that the limit of toxicity of illuminati 
gas on the flower of the carnation is determined by the ethylene 
