62 



foliage in the raspberry is correlated with yellow or lighe colored fruit, 

 and dark fruit is obtained from plants having dark foliage and canes. 

 Eoses, caonas, and asters generally follow the same lines — pale colored 

 foliage indicating light colored flowers In comparing these factors, fully 

 matured leaves should always be examined and even then exceptions will 

 be noted. A correlation is said to exist between the color of the flowers 

 and the seeds of beans, between the color of the roots and stems and the 

 flowers of carnations, between the color of the seed coats and character of 

 plants of peas, etc. There is apparently some correlation between the 

 size of the different organs of plants and possibly between their size and 

 color, but as yet the evidence is not sufficient to formulate definitely. 



0. F. Cook, of this Department, discussed Evolution under Domestica- 

 tion, claiuoing that it is not rational to attribute to environment all of the 

 changes found in plants and animals. 



The Varying Tendency and Individual Prepotency in Garden Vege- 

 tables was tne subject of a paper by W. W. Tracy, of Detroit, Michigan. 

 The author's long experience has enabled him to examine an immense 

 number of specimens of different vegetables and note some of their peculi- 

 arities. These variations will frequently be apparent only from a careful 

 study of a great number of plants. It is stated that different plants of the 

 same natural order tend to vary ab'Ug parallel lines, and variations that 

 are frequently attributed to hybridization are due to ordinary variation. 

 Different natural orders of plants are differently affected by soil, climate, 

 etc. In some cases plants grown from seed from widely different regions 

 did not show any appreciable difference; in others, marked differences 

 may be noted in comparatively slight changes. An example was cited in 

 which a variety of watermelon was grown in Michigan aud in a Gulf 

 State, the seed being from the same source, and the progeny was so simi- 

 lar as to defy any detectable differences. The effect of cultural and cli- 

 matic influences is cumulative. This was illustrated by the simultaneous 

 occurrence in widely separated regions of bush forms of Lima beans, 

 sweet peas, etc. These plants ordinarily grow in climbing forms, but 

 bush forms suddenly appeared in different parts of the United States, 

 seeming to indicate that the influences had been simultaneously working 

 in different regions. Stock produced by an individual grower will vary 

 widely during different seasons. Seed from the same stock, equally well 

 grown under precisely the same conditions, show marked differences in 

 the tendency to adher-e to type in different seasons. Seeds from indi- 

 vidual pedigreed stock plants vary widely in their progeny, and the only 

 way to secure uniformity is to define an ideal, select carefully, and propa- 

 gate carefully, so as to secure a lineal descent of a single typical plant. 



Dr. D. Morris, Imperial Commissioner of Agriculture for the West Indies, 

 read a paper on the Cross Fertilization of the Sugar Cane, in which he 

 described the experiments in the West Indies in the improvement of sugar 

 cane. The subjects for consideration are a greater tonnage of cane, a 

 greater yield of juice, a higher sugar content, and a cane immune to diseases. 

 More than 60 varieties of canes have been imported from all parts 

 erf the world and tested at the various West Indian stations Special at- 

 tention has been paid to bud variation, and a few examples have been 

 found, which were briefly described. Some of these are quite promising 

 as improvements over the older varieties, and so far they tend to come 

 true to color. Nearly all of the bud variations or sports which have been 

 under investigation originated from the ribbon or striped cane. Planting 

 from different parts of cane has not given results of any great value or im- 

 provement, and selection by analysis of the cane juice does not appear to 

 off'er, promise of definite results, nor has selection proved of value when 



