288 



The Journal of Heredity 



"round" the leaf. From these data it 

 will be seen, by comparing the averages 

 of No. 38 with those of No. 33 (the 

 ordinary Connecticut Havana), that 

 the number of leaves is on the whole 

 greater and the average leaf size some- 

 what larger in the Connecticut Havana 

 38. The leaf shape is also considerably 

 better under good growing conditions, 

 though this character is less pronounced 

 than other characters under unfavorable 

 growing conditions. The more signif- 

 icant facts are summarized as averages 

 of eight years in Table H. 



The value of this strain is shown by 

 the fact that in addition to the seed 

 distributed by the station, at least four 

 growers are producing the seed for 

 commercial purposes in one instance on 

 an extensive scale. It is estimated that 

 at least ten thousand acres of this type 

 will be grown in ]]'isconsin during 1919, 

 out of a total acreage of forty to forty-five 

 thousand. This is, then, a case of a new 

 strain of tobacco developed as a result of 

 experimental trial which has been ac- 



cepted and grown commercially . There 

 have been perhaps a number of such 

 failures in tobacco culture leading to 

 the popular opinion among growers and 

 dealers that tobacco cannot or should 

 not be "bred up," but that we should 

 strive to maintain the "original" as 

 closely as practicable, separating our 

 seed by as few generations as possible 

 from the original stock of seed upon 

 which the success of the tobacco in- 

 dustry was founded. The fallacy of 

 such an argument is evident from an 

 agricultural standpoint, although at the 

 same time one cannot argue too strongly 

 against the promiscuous introduction of 

 new strains of tobacco without having 

 first repeatedly tested them out under 

 the commercial conditions for which 

 they are intended. Much progress can 

 be made along the line of improving 

 tobacco by breeding or selection, but 

 it should not be attempted by anyone 

 who is not willing to devote at least 

 eight to ten years of time and work to 

 accomplish it. 



T.VBLE II. — Summarized Data of Most Significant Characters of Connecticut Havana No. 38, 

 Together with Parent and Normal Strains. Averages of eight years. 



Notice to Readers 



Attention is invited to the footnote on the 

 title page, in which it is stated that the Journal 

 is published monthly with the exception of 

 July, August and September. The omission 

 of these three numbers was made necessar>' 

 last year by war conditions, and the fact that 

 the entire editorial staff had entered the 

 military service and left the magazine in Jhc 

 hands of volunteer workers, making the issu- 

 ance of the summer numbers impracticable. 



It is considered advisable to pursue the 

 same policy this year, but the editorial staff 

 has recently been reorganized, and with the 



appearance of the October issue it is expected 

 that the Journ.\l will return to its pre-war 

 basis of twelve issues a year. 



The JouRNAi- relies upon members of the 

 American Genetic Association for interesting 

 inaterial upon subjects falling within its scope. 

 Members are cordially invited to submit 

 articles suitable for publication, especiall}^ 

 those which are illustrated with striking or 

 unusual photographs. It is a particular policy 

 of this publication to interpret the discoveries 

 of modern science by means of photographs. 



