August 9, 1889.] 



SCIENCE. 



parently commonly misapprehended. Its primary value is the de- 

 termination of the vitality of a given sample. This testing, except 

 in rare instances, should be conducted by the grower himself. The 

 proper work for the experiment station is that of determining the 

 best methods and conditions of testing each species and variety : in 

 other words, it seems that the sphere of the stations is to discover 

 and announce laws and rules, rather than to perform the petty 

 tests for the multitude. Merely testing seeds for the purpose of 

 determining how many will grow, is surely not experiment, and 

 the publication of disconnected tests seems to be entirely unprofitable. 

 The endeavor to determine the relative merits and honesty of 

 seedsmen, by means of testing their seeds, is the merest folly. 



There appears to be no necessity for seed-control stations in this 

 country, certainly not for such seeds as fall to the hands of the 

 horticulturist. The control stations of the Old World have suffi- 



cases, duplicate tests were made in an out-door cellar which was 

 used for the storing of nursery stock. 



The conclusions from the tests — seven ^ with beans (Green 

 Flageolet), one with peas (White Garden Marrowfat), one with 

 radishes (Half-Long Early Scarlet), two with turnips (Red-Top 

 Strap Leaf), and four with onions (Giant Yellow Globe Rocca) — 

 were as follows : — / 



1. Different results are obtained from the same sample of seeds 

 under different variations of temperature, of which the daily mean 

 is essentially the same. 



2. Sprouting takes place more quickly under essentially constant 

 temperature of about 74° than under a temperature ordinarily 

 variable, which gives about the same mean. 



ciently exposed the tricks of seedsmen, and have rendered open 

 dishonesty unprofitable. There is now such sharp competition in 

 the seed-business, that seedsmen themselves must exercise every 

 caution in order to demand trade. Improved methods and ap- 

 paratus for harvesting and cleaning are giving us clean samples. 

 The greatest risk in the purchase of seeds is the possibility that 

 inferior strains or varieties may be procured ; but this is a risk 

 which the control station could not assume to govern, inasmuch as 

 the substitution becomes apparent only when the crop is grown 

 The experiment stations may be expected to influence sufficient 

 control in the seed-business, as occasion shall require. 



The tests enumerated in this article have been conducted with 

 the greatest care. Unless otherwise recorded, they have been 

 made in a steam-heated forcing-house. As a rule, they have been 

 made in earth, in shallow earthen seed-pans. These pans are ex- 

 ceedingly convenient, and they afford good drainage. In some 

 cases, lily-pans have been used, but they differ from the seed-pans 

 only in their circular outline and somewhat greater depth. Illus- 

 trations of seed-pans may be seen in Figs. 3 to 7. For sowing 

 seeds at uniform depths, two devices have been used. The simpler 

 of these (Fig. l) is nothing more than a block of half-inch stuff. 



3. Rapidity of sprouting is particularly marked in beans and 

 peas. 



4. As the mean temperature becomes lower, rapidity of sprout- 

 ing becomes slower. 



5. Greater rapidity of sprouting does not appear to be correlated 

 with greater per cent of total sprouting. 



6. Constant temperature, of the degree here mentioned, does not 

 appear to give greater percentages of sprouting: at least, the varia- 

 tion in this respect between the constant and variable temperatures 

 is no greater than that which is usually obtained from tests con- 

 ducted under identical conditions. In the seven tests with beans, 

 however, there is an average gam of 5 per cent in favor of those 

 under constant temperature. 



II. Influences of Different Quantities of Water. 



Mr. W. W. Tracy of Detroit, well known as an expert in the 

 handling and testing of seeds, once said that he rarely obtained the 

 same results from different tests of the same sample, if made in 

 houses under the care of different men. He attributed this varia- 

 tion mostly to the various amounts of water habitually used by the 

 different men. Acting upon this suggestion, a number of very 



two inches wide, of the required length, upon which is nailed a cleat 

 equal in thickness to the depth of sowing. The cleat is pressed 

 into the soil evenly, and the seeds are dropped into the furrow it 

 makes. The other device (Fig. 2) may be called the Tracy 

 planter. It consists of two strips of heavy tin plate nearly three 

 mches wide, hung upon two wire pivots or hinges some two inches 

 long. At their upper edges, and equidistant from either end, the 

 plates are joined by a firm spiral spring, which serves to throw 

 the upper edges apart, and to cause the lower edges to join. This 

 trough is filled with the required number of seeds, and is then 

 inserted into the earth to a given depth, when the fingers push in- 

 ward on the spring, and the trough opens and delivers the seeds. 



ences of Constant and Variable Temperatures. 



The tests here enumerated were made in an incubator of which 

 the temperature was controlled by a galvanic current communicat- 

 ing with clock-work, and in a, steam-heated forcing-house. In 

 the incubator the temperature rarely varied three degrees, while 

 the position of the seed-table in the forcing-house was such that 

 variation sometimes amounted to sixty-five degrees. In some 



careful tests have been made in weighing the amounts of water 

 used. The results have been the most marked of any which have 

 ever come under Professor Bailey's notice in the testing of seeds. 



The tests were all made side by side in a forcing-house, unless 

 otherwise recorded, in earthen pans. The soil, with one excep- 

 tion, was a good quality of light potting earth, containing a good 

 admixture of field-sand. Although the pans were very shallow, 

 extra drainage was given by the use of broken pots. The samples 

 which received the most water were rarely wet enough to drip : in 

 fact, they had no more water than is given in many houses. The 

 pans sparingly watered were dryer than they would be kept in most 

 houses. The 8-inch round lily-pans are 4.V inches deep. The 10- 

 inch seed-pans are 2+ inches deep, and the 12-inch pans 3 inches 

 deep. 



The conclusions drawn from the tests — two with tomatoes 

 (Green Gage), two with cucumbers (Nichols Medium Green), one 

 with lettuce (Boston Market), two with carrots (\'ermont Butter 

 and Early Forcing), one with celery (White Plume), one with 

 turnips (Early Six Weeks), one with pepper (Golden Dawn), two 

 with Lima beans (Large White), and two with Cob(Ea scandens 



