PRINCIPLES OF SEED TESTING 361 



6000 to 15,000 weed seeds were present in i Ib. of what 

 was being sold as red clover, and that in samples of 

 alsike clover as many as 23,556 weed seeds were found 

 in one case, and 49,830 in another. In the former case 

 the sample, having only 72 per cent of pure seed, was 

 priced at $5.25 ( = 2is. lod.) per bushel, or equal to $7.29 

 (z=3os. 4d.) per bushel for the good red clover seed it 

 contained. In the latter case the sample priced at 

 $8.00 ( = 335. 4d.) per bushel, and containing only 48 per 

 cent of good seed, would really cost $16.66 ( = 693. 5d.) 

 per bushel for the alsike. " These (says the Advocate) 

 are startling figures, and though those quoted may be 

 extreme cases, all will readily agree that if the average 

 sample is one-twentieth as bad in this respect, the con- 

 dition is truly alarming." 



7. In the United States of America clover seed tail- 

 ings are sometimes used on the farm, the clean seed 

 being sent to market. Such tailings have been found 

 to contain nearly 272,500 weed seeds per pound. A 

 sample of clover seed offered on the Chicago market 

 about 1898, for 2 cents a pound, contained about 

 338,300 weed seeds per pound, or more than 20 

 millions per bushel. 1 



8. The Bulletin referred to in the last paragraph 

 calls attention to a point worthy of earnest attention, 

 for it is a point frequently overlooked. " Some samples 

 of seed contain such a small amount (of weed seeds) 

 that they would be considered practically pure, but 

 even in these the number of weed seeds in a pound is 

 surprisingly large. In one sample that contained in all 

 only one-fifth of i per cent of spurious seeds, the num- 

 ber of weed seeds per pound averaged about 990. In 

 a bushel of 60 pounds there were, therefore, more than 

 59,000 weed seeds." It is quite clear that very small 



1 U.S. Dept. Agric., Farmers' Bulletin, No. ill, 1900, p. 9. 



