— 366 BOTANICAL GAZETTE [MAY 
to our attention. As we have already pointed out (6), these losses 
generally occur during cold periods in winter. This insures a 
frozen crust, promoting lateral diffusion of the gas from the faulty 
mains; it also prevents ventilation of the greenhouses. So far as 
evidence for the constituents that produce the injury goes, it 
suggests the heavy hydrocarbons, though it is not by any means 
proved that these are responsible for all such injuries. A source of 
greater loss from illuminating gas is injury to shade trees. This 
injury is through the roots, in contrast to the injury in greenhouses. 
We are unable to state as yet what constituents produce the injury, 
though Mr. Harvey’s work in this laboratory indicates that 
ethylene determines the toxic limit of the gas for the roots of Vicia 
Faba. Even so, it is possible that the less volatile materials of the 
gas accumulating in the soil may really be the source of injury, 
the power of accumulation overbalancing the higher toxicity. The 
determination of the constituents producing the injury and the 
tenacity with which they adhere to the soil are of great importance. 
They determine how soon and under what conditions replacing of 
the dead trees by new ones can be carried out. Mr. HARVEY 
is now attempting to answer these questions. The odor-producing 
substances of illuminating gas are retained in the soil with great 
tenacity, but so far as the pea epicoty] is concerned, these substances 
are innocuous, at least in concentrations easily detected by smell. 
Natural gases are generally low in heavy hydrocarbons; in fact, 
those of the Appalachian system bear none so far as chemical tests 
indicate (1); they consist mainly of methane and ethane. This gas 
should be very low in its toxicity to plants. The Baku natural 
gas is said (30) to contain some olefines. 
The few facts established in this field suggest the need of ration- 
alizing various practices in vogue and summarily abolishing others; 
for instance, the practice of burning tobacco stems in greenhouses 
for killing insects. This is a matter of differential poisoning, apply- 
ing a poison that will kill the insect without injuring the plants. 
The processes volatilize nicotine and set free carbon monoxide, 
ethylene, and other gases. So far as we know, it is not certain 
which is the insecticide. If it is nicotine, why not volatilize nico- 
tine from an extract and avoid the deadly plant poison ethylene ? 
