116 BOTANICAL GAZETTE [FEBRUARY 
and the field is abandoned, receiving no further irrigation, which is 
necessary during the growing period of the crop. As a rule, there 
is sufficient moisture in the soil to produce a crop of suckers from 
the old roots. These make a weak growth among the weeds of the 
abandoned fields and produce flowers and seeds. It is this crop 
of suckers which pasnence the seed supply for the new crop of 
tobacco. 
Under these conditions, any form of selection is impossible, 
for the sucker shoots do not show the characteristics of the parent 
plant. When the seed is mature, all forms, good and bad, are 
indiscriminately gathered and resown the following season. 
These methods of obtaining seed and growing posturas bring 
about two results: (1) all types of tobacco that occur in Cuba are 
maintained there by the blanket method of harvesting seed indis- 
criminately from all kinds of plants; and (2) by reason of the traffic 
in posturas and seeds, all types are distributed to all the tobacco- 
growing regions, so that a uniform mixture of types is maintained 
over the island. 
Cultural experiments 
As has been stated, a study of the plants in the field is not suffi- 
cient to disentangle the mixture of types and lead to exact informa- 
tion regarding their constancy. To determine whether these types 
are constant, or whether Cuban tobacco possesses the enormous 
variability usually attributed to it, cultural experiments were 
egun in 1908. During the tobacco season of that year, about 
30 plants were selected which seemed to represent distinct types- 
A careful description, recording all characteristics that might be 
of any value in identifying the types, was written for each plant. 
The plants were staked and labeled and given a number. — Inas- 
-much as the plants had been topped, it was necessary to save seed 
from the suckers which appear at the base of the plant after the 
stem is cut. A number of the plants did not form suckers, so that 
only 14 plants remained. The suckers selected for seed were 
inclosed in manila bags in the usual way, while the others were 
cut as soon as they appeared. In this way, guarded seeds from 
14 isolated, self-fertilized mother plants were obtained. 
