64 



the plant-breediug laboratory. This has included the selection and hybri- 

 dization of oats to secure rust resistance, hardiness, increased yield, to 

 prevent lodging, etc. As yet the results can not be definitely determined. 



In experiments to increase the yield and hardiness of winter oats by se- 

 lection, sowings were made at different dates late into the autumn, and the 

 hardiness of the plants was tested by freezing out during the winter. It 

 was found that the farther south seed oats are produced, the earlier tbe 

 crop when sown at Washington; and there are apparently no varieties of 

 oats which come absolutely true to type description. In experiments in 

 crossing, only 5 to 10 per cent, of successful fertilizations were ordinarily 

 secured. If cool, moist days were chosen for pollination, better results 

 would be obtained, in some casei almost 100 per cent, of the pollinations 

 resulting in the production of fertile seed. The presence of natural crosses 

 was briefly commented upon, and it was stated that Rimpau in his work, 

 covering many years, observed only 4 or 5 cases. This seems to indicate 

 that oats are nearly always self -fertilized. 



In commenting upon this paper, Director Saunders, of the Canada Ex- 

 perimental Farms, called attention to the Canadian work along the line of 

 oats breeding, which has been carried on for about 10 or 12 years, and 

 D. Gr. Fairchild referred to the work being conducted at the experiment 

 station at Svalof, Sweden, along similar lines. 



The subject of breeding Florists' Flowers was treated in papers by E. 

 G. Hill, of Richmond, Indiana, C. W. Ward, of Queens, Long Island, and 

 A. Wintzer, of West Grove, Pennsylvania. In Mr. Hill's paper notes 

 were given on breeding experiments with roses, carnations, and geraniums. 

 Of many thousand hybrids produced, but few of desirable quality were 

 obtained. Not one in a thousand was said to in any way approximate to 

 the value of the ideal which was sought. When the great number of hy- 

 bridizers who are experimenting with roses is considered, the number of 

 valuable new sorts is very meager compared to the amount of labor ex- 

 pended. While rose fertilization is said to be very easy, the most im- 

 portant part of the work is in properly maturing the seed and propa- 

 gating it. A number of successtul hybrid roses were cited as being re- 

 cently introduced into the market. Somewhat similar results were report- 

 ed with carnations, in which out of thousands of seedlings grown every 

 year by hundieds of growers, only a few improved forms are annually ob- 

 tained. In experiments with chrysanthemums the best results have been 

 obtained, it was said, when only double forms were used for parent stocks. 

 The hybridizing of begonias was also commented upon, and the origin of a 

 number of the finest new varieties was indicated. Attention was called to 

 the fact that La France, claimed to be the parent of many new varieties, is 

 absolutely sterile in some countries. 



The result of experiments in crossing pumpkins were described by L. 

 H. Bailey, of Cornell University, in a paper entitled A Medley of Pumpkins. 

 The author began a series of experiments in 1887, which was continued for 

 10 years, to determine the immediate effect of pollen on fruit. After a 

 number of years' investigation, no immediate effect of pollen could be re- 

 cognized on cucurbit fruits. The experiments were continued, however, 

 to see what would be the result of crossing 2 varieties of squash. These 

 Tarieties were hand pollinated, the seed saved, and in the third generation 

 the plants occupied between 8 and 10 acres. Of the product examined, 

 fully 1,500 forms were noted which did not resemble either parent in form 

 or shape. The plants seemed to be almost wholly self-sterile to their own 

 pollen. Ihe seed of one form, designated as Alpha, when planted, gave 

 110 distinct kinds of fruits and inumerable intermediates. The parent 

 stock of this experiment was pedigreed and usually came true to type, but 



