149 



that tbe juice expressed was more dense than ordinarily. It might also 

 be used to illustrate the imperfection of the present method of deter- 

 mining the composition of a cane by the analysis of the juice expressed 

 from it by a mill. Such analyses are always subject to the variations 

 of the degree of extraction by the mill, the dry ness of the cane, etc. Of 

 course the removal of part of the water from the juice on account of the 

 cane being frozen would not often occur, but a loss of water by drying 

 would also have the effect of increasing the density of the juice ex- 

 tracted. It is to be hoped that methods will be perfected that will ad- 

 mit of the proper sampling of the cane itself, and the direct determina- 

 tion of the sugar. 



SIZE OF SORGHUM SEED. 



The size and weight of sorghum seed varies greatly in different varie- 

 ties, and in different individuals of the same variety. Professor Henry 

 found 27,680 seeds to the pound of Wisconsin Amber. Dr. Collier found 

 19,000 in Virginia Amber. In a sample of the Early Amber seed grown 

 at this station there were 20,200 seeds to the pound. In a sample of 

 the New Orange variety there were 21,760 seeds to the pound. In a 

 sample of Donra (non-saccharine) there were 10.480 seeds to the pound. 

 This variety has the largest seed of any grown here. In an average 

 sample of the Red Liberian variety there were 31,400 seeds to the pound. 

 This has the smallest seed of any variety grown here. In a pound of 

 seed of the same variety, selected for large size, there were 21,800 seeds, 

 one- third less than the average sample. 



The vigor of the young sorghum plants in the first weeks of their ex- 

 dstence corresponds closely to the weights of the seeds which produced 

 them. 



It seems evident that more vigorous plants can be procured by select- 

 ing seeds which are above the average size. The Liberian, for instance,, 

 produces very small seeds, and these produce very small and slow-grow- 

 ing plants while they are young, although they eventually produce large 

 and handsome canes. It will be noticed that the sample of larger seeds 

 selected from the Liberian had the same weight as average seeds of 

 other varieties. 



It is to be supposed that these larger seeds would produce more vig- 

 orous plants than the average seeds of that variety. 



Major Hallett found that by selecting the finest grains of wheat he 

 improved the plants and also improved the variety. 



Mr. Wilson separated the largest and the smallest seeds of the Swed- 

 ish turnip; he found that the plants from the largest seeds took the 

 lead and maintained their superiority to the last. 



Director Briem made similar experiments upon sugar beets, as fol- 

 lows:* 



" It is a well-known fact that seeds of different size and weight of any plant will cor- 

 respondingly develop plants of different size and .weight if conditions of life are 



* Wiener Land wirtschaftlioheZei tang, 18^7, No. 99, S. 703. Agric. Science II, 111. 



