cotton, cowpeas, and watermelon was described and photographs and' 

 material shown. As a result of continued selection of resistant varieties,. 

 it was stated, Sea Island cotton is now grown in regions which had been 

 practically abandoned on account of the destruction caused by the wilt. 

 No varieties are wholly resistant, but a number of strains have been found 

 which are to a great degree able to resist the fungus. In general, upland 

 cotton seems less resistant to wilt than Sea Island, and Egyptian varie- 

 ties are more resistant than any of the others. The work so far has been 

 one of selection, since hybridization is not practicable, as it tends to 

 destroy the merchantable character of the fiber. Similar results in the 

 selection of cowpeas and watermelons were cited, and the author believes 

 that many other varieties of plants may yet be found that are resistant 

 to disease. 



W. M. Hays, of the Minnesota Station, presented a paper on Breeding 

 for Intrinsic Qualities. He believed the value of plants and animals 

 annually produced in this country could be readily increased 10 per cent, 

 at an expense of less than 1 per cent. The greatest financial gains would 

 probably be secured by the improvement of a score of plants and about 4 

 species of animals. By carefully growing and testing many thousands of 

 individuals there will frequently be found some one individual of such 

 superior merit as to repay all expense. In any hybridization work a 

 good foundation stock must first be produced upon which to base the new 

 varieties. The importance of working with large numbers of individuals 

 and the value of correlated qualities were pointed out. Durino- the 

 progress of the work various side lines may enter, but these must be 

 held subordinate to the main idea which controls' the experiment. In 

 choosing varieties, often very perplexing problems arise. As a rule 

 crosses should be made between individuals which closely approximate the 

 ideal, and not between those which are too dissimilar. An illustration of 

 the value of using large numbers was given in the experiments in breed- 

 ing wheats. To begin with, 500 plants were examined for loundation 

 stock and tested from 3 to 5 years to see that they came to true seed. After 

 continuing the work this length of time, all were rejected but about 50, 

 which were given a fiel i trial. This work has been continued, large 

 numbers of individuals being constantly grown, and as a result 2 or 3 

 varieties have been found which are intrinsically of great value. In 

 wheat breeding the author thinks that an increased yield, irrespective of 

 distinguishing marks, should be the ideal sought, 



A paper on the Correlation between Different Parts of the Plant in 

 Form, Color, and Other Characteristics was read l)y S. A, Beach, of the 

 New York State Station. By means of a number of specimens he showed 

 the correlation which exists between difi"erent parts of plants. By the 

 proper study of the correlation of form, color, vigor, etc, undesirable ma- 

 terial may be eliminated to a great degree while the seedlings are still 

 young. In this way much valuable time and space may be saved. Nu- 

 merous examples were cited in which this early elimination is possible. 

 Small foliage is said to be usually correlated with small fruit. The text- 

 ure of the leaf and of the fruit are believed to be correlated to some de- 

 gree. JJwarfed seedlings produce poor plants even when given good culti- 

 vation. Attention was called to the importance of considering groups of 

 characteristics. In the case of the peach there seems to be a direct re- 

 lation between the size of the foliage and the size of the fruit, and suffi- 

 cient evidence is believed to be at hand to show that size and color of fo- 

 liage and flowers may be depended upon in predicting the character of 

 fruit. Pale or light colored blossoms are usually associated with small 

 fruit in apples, while deep color is correlated with larger fruit. Pale 



