32 THE NURSERY-MANUAL 



"In the first place it is necessary to get a representative sample by 

 taking seed from top, sides, middle and bottom of sack or bin and to mix 

 thoroughly on a cloth, paper or smooth table. A thin-bladed knife is a 

 convenient tool with which to mix. The person should exercise care not 

 to lose seed or dirt while mixing, as a good test requires careful weighing. 

 When satisfied with the mixing, let the tester divide the sample into two 

 approximately equal parts, discarding one-half and mixing the other, and 

 repeat dividing and mixing until the pile is small enough to analyze. 



"The first step in analysis is to weigh the sample carefully and record 

 the weight. Analysis proper may then begin. This consists of separat- 

 ing the sample into the four or five piles already mentioned. A pair of 

 forceps and a hand lens aid materially, and for small seed, such as grass 

 and clover or alfalfa, are necessary. A needle fastened in a wooden handle 

 or a sharp hat-pin serves to pick out small particles or to separate ad- 

 hering bodies. After the separation is complete, each pile should be 

 weighed, or closely estimated by counting. 



"If possible, the weed seed ought to be identified in order to prevent 

 the introduction of bad weeds new in the locality or not yet started on the 

 farm in question. Identification enables the farmer to know just what 

 he is planting, permitting him to exercise choice as to whether such seed 

 is the best he can find at a reasonable price." 



The complete seed-test 



If a full test is made, it includes an analysis of purity and a 

 record of viability. Following is the form of report recom- 

 mended by Circ. 23 of the Utah Station : 



A report of the test should show the analysis and percentage 

 germination. The following form of report gives the most 

 essential information : 



Kind of seed Variety 



Source Proposed use 



Locality to use it 



Test Per Cent 



Purity 

 Injured seed 

 Other crop seed 



