EXPERIMENT STATION. 103 



Most of them were of good quality as far as vitality is con- 

 cerned. One sample of onion seed, 173, was evidently old and 

 comparatively worthless. 31.0 per cent, sprouted, while 39 per 

 cent, remained hard at the close of the test. One other sample, 

 17], is undoubtedly old seed. 



It will be noticed that a laboratory test can only take into 

 account three of the factors which determine the quality of a 

 sample of seed, viz : its purity, that is, its freedom from seeds of 

 other species of plants, its germinating power, and its weight. 



At present it is not possible in most cases from such a test to 

 decide whether the seed will produce vigorous or feeble plants, 

 or whether it is true to name, in so far as to be free from all 

 other varieties of the same species. 



Appaeatus for Tksting the Vitality. of Seeds. 



During the year the apparatus here described has been tested 

 and found to give perfectly satisfactory results with onion and 

 lettuce seed. Other species of seeds have not as yet been tried 

 in it. 



The apparatus consists of a pan of copper or galvanized iron 

 two and a half feet long, ten and a half inches wide and one and 

 three-quarters inches deep. At one end is a small horizontal tube 

 let into the side near the bottom, which may be closed with a cork. 

 The pan has a slightly arched cover, two feet and four inches 

 long, ten inches wide, with a rim three inches deep. It is pro- 

 vided with a handle and has two three-quarter inch orifices on 

 top through which a thermometer can be introduced if desired, 

 which also secures sufficient ventilation for small seed. The tiles 

 which hold the seed to be tested are made of a very light* and 

 coarsely-porous earthen ware which absorbs water almost as rap- 

 idly and abundantly as a sponge. This material is manufactured 

 by S. L. Pewtress & Co., of New Haven, as a filtering medium, 

 and is quite unlike any other earthenware that has come under 

 our notice. Each tile is nine inches, by eight and one-half, by one 

 and one-half. The upper surface is grooved, by help of a broad 

 file, so as to form seven channels or beds, about one-quarter of an 

 inch apart, running the length of the tile, each seven-eighths of an 

 inch wide and three-sixteenths of an inch deep. 



* One of them weighs, when dry, 2 pounds 1 1 ounces. 



