1 32 BOARD OF AGRICULTURE. [Jan., 



tobacco by seed selection, and probably by hybridization, is one 

 of the most important means of increasing the value of the crop, 

 and without additional cost of production. The general results 

 of the experiments conducted by the Department of Agricul- 

 ture with the breeding of different varieties of tobacco in the 

 Connecticut Valley last season will be presented here, in order 

 to offer some practical suggestions to growers for the improve- 

 ment of their crop by breeding and seed selection. 



CHANGE OF SEED. 



There is a general impression that a change of seed from 

 one region to another is necessary to prevent the " running 

 out " of the variety. In other words, many growers believe 

 that a variety of tobacco will deteriorate in quality if grown 

 on the same farm or in the same locality for several years. A 

 similar belief exists among the growers of other crops, such as 

 corn and cotton, and many of those farmers follow the practice 

 of obtaining fresh seed from some distant locality year after year. 

 In corn and cotton it has been found by careful tests that when 

 a variety has become adapted to certain soil and climatic con- 

 ditions the yield may be increased by seed selection and breed- 

 ing, and, instead of deteriorating by reason of continued grow- 

 ing in the same locality, the value of the crop can be increased. 

 In order to illustrate the effect of a change of soil and climatic 

 conditions upon tobacco the experience of the growers of im- 

 ported Sumatra and Cuban plants in this region will be re- 

 viewed. In the seasons of 1902 and 1903 tobacco seed from 

 Cuba and Sumatra was extensively used on the plantations 

 growing tobacco under shade. The crops produced from this 

 seed lacked uniformity to such an extent that several distinct 

 types of plants could be found in every field, many of these 

 types bearing no resemblance whatever to the Cuban or Su- 

 matra varieties. The Cuban seed produced a large proportion 

 of small, narrow leaved plants, the so-called " freaks," in some 

 cases constituting about one-third of the entire crop. The 

 Sumatra seed produced many long and pointed leaved types, 

 which were practically worthless for any purpose, and were 

 probably reversions to some early unimproved type of tobacco. 

 This breaking up of type and lack of uniformity has been 

 observed to follow the use of imported seed in other tobacco- 

 growing sections of the United States, particularly when seed 



