THE CORRELATION BETWEEN LENGTH OF FLOWERING-STALK 



AND NUMBER OF FLOWERS PER INFLORESCENCE IN 



NOTHOSCORDUM AND ALLIUM. 



BY J. ARTHUR HARRIS. 



Several years ago DeVries in outlining some of the prob- 

 lems of chief importance for future investigation in evolu- 

 tionary work, designated correlation as one of the most 

 important.* In another place in the same volume he dis- 

 cusses data of his own for the relationship between the 

 diameter of the stem and the length of the fruit m certam 

 of his evening primroses.! 



Notwithstanding DeVries's emphasis of the importance of 

 quantitative studies of correlation, comparatively little has 

 been done by botanists. His suggestion concerning a rela- 

 tionship between the size of the individual and the measur- 

 able characters of the fruit called my attention to the 

 desirability of a series of comparative studies of the inter- 

 relationship between the vegetative development of organs 

 of the individual and the number or fertility of the flowers 

 or fruits which it produces. 



As material for a first determination of the interdepend- 

 ence of length of flowering-stalk and number of flowers per 

 inflorescence I shall use series of Nothoscordum and Allium. 

 Data from the inflorescences of several other species are 



being examined. . 



In the spring of 1906, a collection of the flowering-stalks 

 of Nothoscordum striatum was made just south of the railway 

 station at Mcramec Highlands, near St. Louis, Mo. Here the 

 brow of the hill slopes in two directions, to the southeast 

 and to the southwest. The ground was more or less covered 

 with old dead or leafless vegetation consisting chiefly of 

 sumach bushes, large tufts of grass and weeds, and low grass. 

 While no quantitative determination was made, it seemed to 

 me that the old vegetation on the southwest slope was much 



♦ DeVries, H. Die Mutationstheorie. 1 : 113. 1901. 

 t DeVries, I c. 381-383. 



(105) 



