86 PASTURE PLANTS AND PASTURES OF N.Z. 



seeds commonly vary by as much as 50 per cent, in the ger- 

 mination capacity, it is clear that this is often the ruling 

 factor in the value of the seed, and appearances count for 

 but little. It cannot be too often emphasised that in sowing 

 grasses to remain down for six to twenty years everything 

 should be done to secure a good start, and a small extra 

 initial outlay is justified a hundred times over if bare patches 

 and an indifferent herbage are thereby avoided. A farmer 

 will studiously avoid sowing shrivelled wheat, or turnip 

 seed reputed to be old it is very much more important that 

 he should avoid graSs seed with a low germination. 



To test seeds is the simplest of matters. Take a square 

 of cloth or of blotting paper, damp it thoroughly, and lay it 

 on a plate. On this place about 100 seeds, spreading them 

 out slightly. On these place another piece of damp blotting 

 paper or flannel, and cover the whole with an inverted plate 

 to keep the moisture in. The temperature of an ordinary 

 living room about 60 degrees F. is best for germination, 

 and during the trial the upper plate should be removed 

 for a second or two each day to allow free admission of 

 air, and the escape of foul gases generated when the plates 

 fit too closely. Certain grasses germinate better in the 

 light than in the dark, for example Poa pratensis, and for 

 these the following plan is perhaps the easiest : Take a 

 piece of porous roofing tile, lay it on the flat, and immerse it 

 for half its depth in a dish of water. On the tile spread 

 out the seeds which will then need no further attention than 

 to keep the water in the dish replenished as it evaporates. 

 This plan is really easier than that of the blotting paper, 

 if only a piece of porous earthenware is available. 



The above methods of seed testing are quite good enough 

 for common use. The best results, however, are obtained 

 when blotting paper is used between glass dishes, and this 

 method must be adopted for accurate determinations. The 



