72 PLANT BREEDING. 



5. Discard 00 to 95 per cent of the jDlants, and take full notes on 

 the remaining plants on a blank form like the following: 



FLAX. 



Nursery No _..... . .. . _.-. . 181 



Type . - ... '- - - . . Fiber. 



Date when ripe .... . - . . . _ . Aug. 15 



Height . . - - . inches . - 40 



Number of tillers - - _ - - - . 1 



Number of branches . .- .. 5 



Number of bolls -.. .-.- . . 6 



Evenness of ripening ... _ _ _ per cent . . 90 



Size of seeds ....do 97 



Grale of seed ... . . do 95 



Gross yield of seed - . . ..grams.. 32.5 



Net yield of seed ..- _ ...do... 16.4 



6. From among each of the best centgeners of each tj^pe choose a 

 few of the best plants for mothers of centgeners the second j^ear, and 

 continue j'ear after year recording, compiling, and averaging the facts. 



7. When any stock has shown a superior habit of growth, yield, and 

 quality, multiply it rapidly and test it in the field, and, if a fiber kind, 

 test it also in the factory or laboratorj^ for yield and quality of fiber 

 in comjiarison with standard varieties. Laboratory methods are being 

 developed for testing the strength and fineness of the fiber and for 

 determining its percentage of the croj), so as to determine the jdeld of 

 fiber. 



8. It may not be wise to limit the parentage of a new flax to a single 

 mother plant, as this is an open-pollinated species, but the seeds fx'om 

 three or more centgeners may be mixed together in attempting to 

 make a new variety. 



9. Stocks once started in the nurserj^ will serv^ as excellent parents 

 to use in creating new qualities by hj'bridizing. 



10. Hybrid stocks may be grown in l)ulk for a few years, that varia- 

 tion may fullj^ develop, the selection then being carried on as under 

 paragraphs 2 to 8, inclusive. 



11. Two varieties may be hybridized by mixing the seed and sowing 

 for two or three years in a plot or field, from which superior mother 

 plants may then be chosen, as already described. 



12. New hybrid varieties should not be distributed to growers until 

 they have been tested several years in uniform test plots. 



O 



