58 BOTANICAL GAZETTE [JANUARY 
that the effect of the carbon black is not due to any stimulating sub- 
stance which it carries into the medium. Other experiments have 
shown that no acceleration of growth is obtained with the addition 
TABLE I 
DATA FOR EXPERIMENTS I TO Iv. 
ExPerIMENT I | Expermwent II | Experiment III) Experment IV 
24 PLANTS 24 PLANTS PLANTS a 
GROWN I9 DAYS | GROWN 19 DAYS | GROWN 17 DAYS | GROWN 13 DAYS 
MEDIUM 
Total 
Total Per | Total Per Bit Per Total Per 
trans- cent. trans- cent. cent. trans- cent. 
piration| increase| piration| increase dav er z increase| piration| increase 
cultur’s) 
anresire Bon extract:.% |. 84:1. cc's ;. }{2) 33 tne 51 Preis, 
o.+tri-calc. phosphate..} 182 | 18.1 i tele ieee : + ates spe tans 
Decicaiiis carbonate..} 191 | 24.0 (2) 8 TAG Al hes corns eee 
Do.+ ferric hydrate. ..... 184 | 19.5 Bae 233.3 | 187 |266.6 |(2) 176} 158.8 
Do. +-sitminwin Wydiate..6 2 487: eta ian be | ek Pe op eee ee 
’ Do.+carbon black ...... ..ee |... | (4) 69/109.0 | 154 [201.9 | (3)139] 104-4 
Do.+magnes. carbonate..| .... | ....-|(5) 73|121.2 ced i Baers ey ones er 
Do.+ barium carbonate...| .... | .... | (6) 78/136.3 aie wee 
Do.+ quartz flour........ pa ee ear es heel Sete 87 Fors 
Do.+ ferric hydrate, car- 
crt and calcium 
etic Ap a ors ae See Wee Py crag cre tk mr en deren es bP pee 
of iron salts to the extract of poor Leonardtown loam. Therefore, 
a slight increase in dissolved iron cannot be the cause of the accelera- 
tion noted in the case of this hydrate. It seemed probable that ferric 
hydrate and carbon black had their effect through an active removal, 
perhaps by mere adsorption, of some injurious substance occurring 
in the culture medium. Such a substance might have been in the 
soil extract originally, or might be produced by the plant roots, or 
both suppositions might be true together. The third alternative 
proved to be the correct one. 
To obtain evidence in this regard, Experiment V was carried out. 
Four different soil extracts were shaken with carbon black, filtered, 
and then used as culture media, comparison being made with controls 
in untreated extracts. The four soils-were of two types, a good and 
a poor soil of the Cecil clay type, and a good and a poor of the Leonard- 
town loam type, the former from Statesville, N. C., and the latter 
from Leonardtown, Md. Chemical analysis both of the aqueous 
extracts and of the solution obtained by digestion with hydrochloric 
