273 



(23) The latitude in wliidi seeds are srown may determine their behavior in 

 germination. 



(24) Northern-grown ('(jrn appears to germinate quicker than Southern-grown 

 corn. It is to be expected,, from our Icnowledge of the variation of plants in 

 reference to latitude, that seeds of most species will give similar results. 



(25) Variation in results of seed tests may be due to the apparatus in which 

 the test is mad(>. 



(26) Ap))aratus in wliich the seeds are exposed to light are to be distrusted. 



(27) Apparatus which afford no protection to the seeds other than a simple 

 layer of cloth, paper, board, or similar co^er, are usually unsafe, from the fact 

 that they allow of too great extremes in amounts of moistur(\ 



(28) The so-called Geneva tester appears to give better results of sprouting 

 than tests made in soil, probably from the fact that moisture and temperatiire 

 are less variable than in the soil tests. 



(29) To study gerniinaticm to its completion, tests iinist be made in soil. 



(30) Tests made in-doors are more reliable than those made in the field. 



2. Results commonly vary between tests made under apparently identical con- 

 ditions, even with selected seeds. Therefore, 



(31) One test can not be accepted as a true measure of any sam])le of seeds. 



3. The results of actual ordinary planting in the field can not be considered a 

 true measure of the viability or value of any sample. 



4. Rapidity of sproutings. iniloss under identical conditions, is not a true 

 measure of vitality or vigor of seeds. 



5. There appears to be no pernicious adultei-ation of garden seeds in this 

 oountr.v. and as a rule there are no hurtful imimrities. 



In the ordinary farmer's garden seed testing is perhaps of little or no value, 

 but to the market gardener, who plants considerable areas to special crops, and 

 to the seedsman, it is highly profitable. It is possible that in some cases the 

 character of the crop can be prognosticated with some degree of certainty from 

 behavior of plants in germination, wholly aside from percentages of sprouting. 

 The studies of experts in this country arfd Germany indicate that when accu- 

 rate information is desired as to the value of seeds the seed test should present 

 at least the following data : Name of variety, where grown, when grown, how 

 keiit, per cent by weight of foreign matter, per cent by weight of apparently 

 good seeds, nature of foreign material, weight of seeds, manner of testing, num- 

 ber tested, average and extreme temperatures during trial, first germinations in 

 hours, last germinations in hours, i»er cent by number germinated, per cent un- 

 sprouted but sound at end of trial, date of test, estimate of agricultural value. 



BULLETIN No. 8, AUGUST, 1889. 

 The effect of different rations on fattening lambs. I. P. Rob- 

 erts, M. Agr., and H. H. Wing, B. Agr. (pp. 75-8(>). — "These 

 experiments were, in the main, a continnation of thone carried on at 

 this Station one year ago, and reported in Bulletin Xo. 11, and very 

 nearly the same foods were used, none of them being out of the reach 

 of the general mass of farmers." The period of feeding lasted five 

 full months, from November 2^. 1888. to April 25. 1880. Twelve 

 lambs in four lots of three each were used, but for reasons stated in 

 the bulletin the computations of results are based upon the two heavi- 

 est lambs in each lot. The lots were numbered respectively III, IV, 

 Y. and VI. Lot III received what may be called a '' carbonaceous " 

 ration, consisting of timothy hay, whole corn, and roots; lot IV, a 

 nitrogenous ration, consisting of clover hay, roots, wheat bran, and 



