64 HORTICULTURAL SOCIETY OF NEW YORK. 



to be planted, and surely with the parents, or stock plants, from which he 

 expects to take the seed, cions, buds or cuttings for propagation. Correlation 

 of parts or characteristics in plants may be of use not only in selecting seed- 

 lings, but also in choosing the parents, and even in choosing the seeds which 

 are to be used in breeding. Notice a few illustrations. 



Geschwind 1 finds correlation between the structure and the sugar con- 

 tent of the beet-root. He states that, as a rule, high sugar content is asso- 

 ciated with a small amount of woody tissue, and recommends that beets be 

 selected for breeding which have but small amount of woody tissue, as shown 

 by cross-section of the top. 



Herislow 2 states that McNab finds that in breeding rhododendrons the 

 best dwarf varieties are obtained by using pollen taken from the smaller 

 stamens. 



Swingle :! states that in Europe certain plant breeders who had long been 

 engaged in breeding grain for the increase of the percentage of protein 

 found recently that a high nitrogen content of the grain is correlated with 

 blue stemmed plants, and since making this discovery have been enabled to 

 make more progress in three years in increasing the nitrogen content of the 

 grain by plant breeding than they had in many previous years' effort toward 

 the same object. 



Henslow 4 says that very dark crimson zonal geraniums are so nearly 

 self sterile as to make seed raising difficult, the sterility being in proportion 

 to the depth of color, which is correlated with proterandry. Paler varieties 

 are more nearly homogamous and are very self-fertile. 



Tinker 5 states that male vines of Vitis bicolor, MX. at all ages, have 

 leaves more lobed or divided than pistillate vines of the same species, and 

 that this distinction is discernible in seedlings when they have put forth 

 the sixth leaf. In his work in breeding grapes he finds it practicable to dis- 

 card male bicolor seedlings when the sixth leaf is formed. 



Debruyker 6 has shown that correlation in length exists between the culm 

 and the head and the upper internode and the head of the rye plant. He did 

 not find, however, that heredity had any apparent influence on these features. 

 Many instances exist of correlation between different parts of a plant in 

 size, but not enough observations have been made to permit of general state- 

 ments as to the full significance of this character in any particular class of 

 plants. De Vries 7 finds that there is a relation between the vigor of the 

 plant in Oenothera Lamarckiana and the size, *. e., length and thickness, of 

 the fruit. The larger and more vigorous the plant, the longer and thicker 

 the fruit; the shorter the fruit, the weaker and more slender the plant. 



In the vineyard of E. C. Gillett, Penn Yan, N. Y., is a vine of the Con- 



1 Geschwind, L., Rev. Gen. Chim. Appl., 3 (1900). No. 12. Cited in Exp. Sta. Rec. 

 XIII: 526. 



2 Garden LX (1901) ; 228. 



3 Swingle, W. T., Discussion on Plant Breeding at the New Haven meeting of A. 

 A. A. C. E. S., Nov., 1900. 



4 Henslow, G., Jour. Roy. Hort. Soc. XXIV: 85. 



B Tinker, Dr. G. L., in personal communication to the writer, 1902. 



"Debruyker, C., cited in Exp. Sta. Rec, XIII (1901): 241. 



7 De Vries, Hugo, Die Mutationstheorie. Erster Band (1901): 377. 



