:>f weed seeds 



SEEDAGE 



jected; also if 1 per cent o 

 found. 



The reference collection of seeds should be kept in 

 neatly labeled glass bottles, without necks, tightly 

 stoppered and systematically arranged in shallow paste- 

 board boxes ( see Fig. 2295 ) . A convenient size for these 

 bottles is 2 in. long by 3-5 in. in diameter. A tray 

 holding 100 of such bottles should fit into an ordinary 

 herbarium case. If the collection is large, a card index 

 will be of great assistance in finding the specimens. 



Germination Tests, — The seeds used in germination 

 tests must be taken indiscriminately from pure seed 

 which has been thoroughly mixed for that purpose. 

 The selection of plump, nice-looking .seeds for these 

 tests, as frequently practiced, impairs the authenticity 

 of the result. 



Tests may be conducted in the laboratory between 

 damp cloths or bluttcrs, or in porous saucers, or in sand 

 or soil in n -r. ■ imi.u . S.i-ds which are known to ger- 

 "uUl be tested in a greenhouse 

 • ry. The same is true of any 

 '-' < Mniitions of germination are not 



SEEDAGE 



1643 



minate \ 



species 

 well. 



adti 



While damp blotters serve as the best .substratum 

 under ordinary circumstances, and especially where a 

 large number of tests are to be made, they do not 

 answer as well for fine, slow-germinating seeds like 

 tobacco and June grass, and many flower-seeds, owing 

 to the fact that the blotters sometimes adhere too closely 

 to permit the proper circulation of air. This may be 

 remedied to a certain extent by placing narrow strips of 

 glass between the folds, but main reliance in such cases 

 should be placed upon soil tests. 



All tests are to be made in duplicate, using two lots 

 of 100 seeds each of peas, beans, corn, cucurbits and 

 others of a similar size, and 200 seeds of clover, cab- 

 bage, lettuce, etc. The more seeds taken for test the 

 less the chance of error. However, 5 per cent to 10 per 

 cent of variation may be expected between the two lots 

 of seed, even though they might have been taken from 

 the same plant. In the case of a greater variation than 

 10 per cent the test should be repeated. Seeds upon 

 which moulds form quickly are likely to be old stock. 



The seeds should be inspected daily, a note being 

 made of those having sprouted, which are then thrown 

 out. In testing seeds of the pea fam- 

 ily (Leguminosse) one-third of those 

 remaining hard and fresh at the close 

 of the test are usually counted as hav- 

 ing sprouted. The average of the 

 duplicate tests is to be taken as the 

 percentage of vitality. Averages 

 should not be made, however, between 

 results obtained by different methods, 

 -uch as blotters and soil. 



Laboratory tests are preferably made 

 iietween damp blotters placed in a 

 metal chamber heated by gas, the heat 

 being controlled by athermo-regulator. 

 The blotters must be free from soluble 

 chemicals. Blue blotters will be found 

 less trying to the eye than white. The 

 irerminating chamber may be of any 

 form which allows proper control of 

 the conditions of light, heat, air and 

 moisture. The standard chamber 

 adopted by the association of Ameri- 

 can Agricultural Colleges and Experi- 

 ment Stations was designed by the 

 ^"3"^"^' writer, and serves equally well for 

 Aeric.'for small bacteriological purposes or experi- 

 seeds ments in plant physiology as for seed 



testing (see Pig. 229())- 

 It IS made of 20 ounce corrugated copper and is 2 feet 

 long 18 inches deep and 2 feet high outside meisure 

 ments The outside except the bottom, is covered with 

 two layers of felt, each K inch thick 



K water spice is ifForded b's the double wills which 



Battle simila 



may be used for the insertion of a thermometer, if de- 

 sired. Owing, however, to the influence which the ex- 

 ternal atmosphere exerts upon thermometers whose 

 tubes are partly exposed, provision has been made for 

 holding two thermometers in a horizontal position, one 

 on the inside of each panel of the door to the chamber, 

 by means of hooks of stout copperwire (Fig. 2297, a, a). 

 The door is made in 2 panels, each consisting of 2 

 plates of thick glass set about K in. apart in a copper 

 frame, which is covered inside with felt. The inside 

 margin of the door is provided with a projection (Fig. 

 2297, c) which fits snugly into a felt- lined groove (Fig. 

 2297, 6), extending around the front side of the cham- 

 ber. The door is 3 in. shorter than the front of the 

 chamber, the remaining space being closed with copper 

 and provided with a ventilator (Fig. 2296, h), which per- 



Used bvtl 

 and Ameru 



jitket 

 into chamber d 

 burner f gas 

 door slides 



mits the exit of carbon dioxid, and can be closed tightly 

 with a slide. Perfect closing of the door is further ef- 

 fected by a copper slide extending along the front mar- 

 gin, which catches flrmly at the top and bottom of the 

 chamber (Fig. 2297, d,d). This device, together with 

 the groove and Its corresponding projection, are adapted 

 from the Rohrbeck bacteriological <'lKinitKT. The out- 

 side door is furnished with a fraiiic into which slide 

 two plates of galvanized iron pniiiti d dead black inside 

 and covered with f.-lt 1 V\kr. 'J2:"',. i , j 1 . By this arrange- 

 ment the intcri.iT' oi' rlic rli.tnilirr may be kept dark or 

 exposed to In , , ' : :: >. .1. mi ,, -half may be dark 

 and the rest ; . ! .n. In ions remaining the 



same. By rai-iti:^ iIm-m 

 read without oiK-niui^ tli 



be 

 Glass plates of various 

 colors may be substituted for the slides, if the effects 

 of different rays of light on plant-growth are to be 

 studied. 



Seven movable shelves, placed 2M in. apart, are held 

 in place by copper ledges H inch wide. These shelves 

 are made of brass rods IH in. apart, and each one is 

 capable of holding up 60 pounds weight. The tempera- 

 ture is controlled by a low-temperature thermo-regulator 

 (Fig. 2296, b). A very low aii.l .■.|iial.l.> llame is secured 

 with a microbunsen Imimr 1 I'i^'. 'IVM't. e). One of the 

 openings into the war.r jacK. f i I'iu'. '1'1W\, a) is 2 in. in 

 diameter to admit a Koiix tlimuo-ri'gulator, if a very 



