~ 
1913] KNIGHT & CROCKER—TOXICITY OF SMOKE 355 
OCs and he 3-4.5cm. long; swollen portion 1-1.5 cm. long, 
wih. declination of.75°—90 
4. 50CC.; epicotyls 3-4 cm. long; swollen portion o.5-0.75 cm. long, 
with little or no declination. 
Washed smoke from tightly rolled paper cigarette 
occ; a 3-4.5 cm. long; swollen portion 1-1. 5 cm. long, with 
declination of 7 
6; 20 Ce.; epicotyl & 5 cm. long; swollen portion 0.75-1.5 cm. long, 
with ane ta of 75°- 
TREC: ee Chay 3-4 cm. long; swollen portion o.5-0.75 cm. long, 
Ww ith little or no declination. 
It appears from these experiments that the tightly rolled 
cigarette gives slightly the more toxic smoke, though not markedly 
Fic. 2.—Responses to paper smoke: a, check; 5, response to 10 cc. of smoke from 
loosely soiled paper cigarette in 10-liter chamber; ¢, response to 20 cc. of smoke. 
so. In both sorts of smoke 10-20 cc. in 10 liters gives NELJUBOW’S 
third response, while 50 cc. in ro liters gives his second response. It 
is apparent that considerable variation in concentration is necessary 
before a noticeable difference in response is shown. Even doubling 
the concentration may modify the response only slightly. This 
condition holds for ethylene, carbon monoxide, and a number of 
other gases, as our later paper will show. On the other hand, with 
pyridine, ethyl and propyl alcohol, acetylene, and others, rather 
slight variation in concentration is evidenced by very noticeable 
differences in the response. A series of photographs will give a 
more vivid idea of the response to smoke. Figs. 2 and 3 show 
the response to various concentrations of paper smoke. 
