54 BOARD OF AGRICULTURE. [Pub. Doc. 



This suggests the practicability of the gardener himself 

 growing the seed he needs. Can it be done, and how ? I 

 would answer that I do not believe that a gardener can profit- 

 ably grow a general collection of seeds, but I do think that 

 in many cases he can grow seed of one variety of the species 

 of which he makes a specialty, and produce seed far superior 

 to any it is practical for the seedsmen to furnish him; for 

 not only does he know just what he wants, but he can give 

 more care to selection than it is practical for the seedsmen 

 to use with any but his stock seed. Again, in many cases 

 seed produced in any location will give better results in that 

 location than seed grown elsewhere, and, as the seedsmen get 

 their stocks grown in the section where because of local cli- 

 matic soil or industrial conditions it can be produced at least 

 cost, local-grown seed is superior; and so, though it may be 

 something of a repetition, I want to speak of how it is pos- 

 sible for the gardener to grow seed most likely to uniformly 

 develop into plants of desired type. First, he must form a 

 very clear and exact conception of precisely the plant wanted, 

 and this should be very fully and accurately recorded, and, 

 if possible, the records supplemented by photographs. I re- 

 gard this exact description and definition of the desired type 

 as very important, — indeed, as the very foundation stone 

 of all successful seed growing. Without it one is very apt 

 to vary more or less in the type of his selections even the 

 first year, and certain to do so in subsequent ones ; and varia- 

 tion of type in the stock seed plants, even if the variation is 

 in itself unobjectionable or even desirable, is fatal to uni- 

 formity of type in the market crop. Having selected a few 

 plants of the desired type, the seed from each of these should 

 be sowed separately and a sample of it planted, in order to 

 select those plants which show the greatest prepotency. If 

 these growing tests of seed from the different selected plants 

 are carefully studied with our description of type in hand, 

 it will often be found that there is a great difference in the 

 uniformity of adherence to the desired type, and often seed 

 from what we considered the very best plant will develop 

 into plants showing the greatest variation of any; and, if so, 

 that lot of seed should be ruthlessly rejected, even if it con- 



