84 



and possibly a superior variety. Sixty of these were considered 

 worthy of further trial and were planted last spring. Three of these 

 are now considered very promising new and distinct varieties, the size 

 and the uniformity of the canes being considered in connection with 

 the average analyses of the juice in 1890 and in 1891. Twelve crosses 

 were also continued, which had given good results in 1888, 1889, and 

 1890. 



The work which has been done in the last four years with varieties 

 at this station shows that sorghum is not the fickle plant it is generally 

 supposed to be. Few of the leading varieties have varied in their mean 

 analysis for any year 1 per cent above or below their general average 

 for four years. The varieties having high percentages of sugar have, 

 when in fair conditions, always retained their high percentage of sugar. 

 The varieties having low glucose have retained that quality, and the 

 varieties having high glucose still retain it. The medium varieties 

 have always beeen medium varieties. 



In continuing the experimental work with varieties of sorghum, it is 

 recommended that six of the best varieties be selected and studied, 

 more particularly with reference to comparative values. The work at 

 this station, so far, has been mainly a search for better varieties among 

 a multitude of varieties. It is not recommended that work on new va- 

 rieties should be discontinued, for the five varieties which have given 

 best results all either originated at this station or were introduced by 

 the station within the last four years, and the fact that the variety 

 which has given the best results of all the varieties in two seasons was 

 received from a foreign country in 1890, is sufficient evidence of the 

 utility of this work. 



It is also recommended that more special attention be given to the 

 improvement of the selected varieties by seed selection. The selection 

 of seed which has been made so far has been more in the line of se- 

 curing pure seed by growing canes from selected seed heads. To show 

 concisely what is desired, the results obtained by Prof. Maercker at 

 Halle, Germany, in improving the sugar beet by selection of seed in 

 ten years (1880-1889) are given here: 



Summary giving average sugar content of beets grown at Halle Station. 



It is obvious that to arrive at such figures, correctly representing a 

 corresponding improvement in varieties of sorghum, will require special 

 attention to this line of work. 



