August 9, 1889.] 



SCIENCE. 



91 



same nature is furnished from the tests here recorded. Several 

 tests were made in which the seeds were sown upon the surface of 

 soil in pots or pans ; the pots, unless otherwise mentioned, being 

 plunged in sphagnum moss to keep the soil moist. Over the top 

 of the pot or pan was placed a pane of glass, or a close-fitting iron 

 ■saucer or a board. 



The conclusions from these tests — one with Papaver rhceas 

 (English poppy), one with larkspur (Dwarf Rocket), one with 

 Adonis astivalis, and one with radishes (Early Scarlet Globe) — 

 were as follows : — 



1. Ver)' great differences in results may sometimes be expected 

 between samples exposed to light during the process of sprouting 

 and those kept in darkness. 



2. When such differences occur, they indicate that light retards 

 or even wholly prevents germination. 



3. In some species this influence o( light is greatly marked, while 

 in others it is not apparent. 



4. It is apparent that those apparatus which test seeds by hold- 

 ing them on a porous plat^ above water are to be looked upon with 

 ■distrust, unless provided with an opaque covering ; and even then 

 they may prove unsatisfactory, as the experience with the larkspur- 

 seeds indicates that best sproutings follow planting in the soil. 



VI. Weight of Seed in Relation to Sprouting. 

 Many experiments have been conducted this year upon the rela- 

 tion of weight of seed to germination, but the figures are too nu- 

 merous to be recorded here. The general results of the tests may 

 be indicated, however. Most of the work recorded in Sections VI. 

 and VII. was performed under the direction of Professor Bailey, by 



VII. Color of Seed in Relation to Sprouting. 



Color may be assumed to indicate, in most cases, some vital 

 character of the seed, as determined by various causes. In one 

 species, or even in one individual sample, it may indicate a different 

 character from what the same color does in another species or sam- 

 ple. It may indicate degree of maturity, method of curing, age of 

 seed, or other peculiarity. It is to be expected, therefore, that color 

 may sometimes designate more or less accurately the germinative 

 vitality of the seed. It follows, however, that no general law of 

 relation of color to germination can be announced : every species, 

 and sometimes every sample, must be investigated for the law 

 which governs itself. Many tests in this direction have been made, 

 but one example will show something of the extent of variation in 

 seeds of different colors. 



With the bean (Green Flageolet), sproutings were most rapid, 

 and higher in total per cent in the green-colored samples. This 

 test was twice repeated with similarly marked results. The same 

 variety from the Department of Agriculture gave opposite results, 

 however. 



Fig. 6 shows tests of white and green colored Lima beans, sown 

 at the same time. The green-colored seeds are ahead. • 



Four tests with morning-glories (both Convolviclus major and C. 

 minor) gave results uniformly in favor of white seeds as contrasted 

 with black ones in the same sample. 



From a considerable study of the importance of color in relation 

 to germination, the following conclusions have been drawn : i. 

 Seeds which differ widely in color in any sample frequently give 

 different results under test ; 2. This variation in results may lie in 



Heavy Seeds. FiG 



Mr. B. R. Wakeman, of the class of 1889, in preparation of a 

 thesis for graduation. 



Of itself, per se, weight appears to exercise no influence upon 

 germination, but it is often a tolerably accurate measure of viabil- 

 ity as determined by various causes. Broadly stated, it may be 

 said that comparative lightness in a seed indicates arrested growth, 

 and consequent lowness of germinative vitality. 



Fig. 5, from a photograph, illustrates a test with radish-seeds, 

 in which the differences were marked. 



It is often true that over-ripe seeds germinate more slowly, and 

 give lower total results, than others, and this over-ripeness is some- 

 times indicated by additional weight. It is to be expected, there- 

 fore, that in some instances best results in germination come from 

 the seeds of lighter weight. 



The conclusions from a number of tests — two with cabbages 

 (one Red Dutch and one Flat Dutch), one with radishes (Early 

 Scarlet Globe), one with beans (Improved Green Flageolet), and 

 one with Lathyrus saiivus (gesse), and others — were as fol- 

 lows : — 



1. Variations in results of testing, both as regards rapidity of 

 sprouting and the total amount, may be expected between seeds of 

 different weights in the sample. 



2. This variation is much greater in some species than in others. 

 In these tests the variation was particularly marked in cabbage, 

 radish, sweet pea, bean, gesse {Laikyrus saiivus), burnet {Poterium 

 sanguisorba), martynia, orach. 



3. As a rule, the heaviest seeds in any sample give earliest and 

 highest results. 



4. In some cases the lightest seeds in the sample give earliest 

 and highest results, apparently because the heaviest seeds, with 

 which they are compared, are over-ripe ; or in some instances un- 

 der-maturity may result in earlier germinations, and such seeds are 

 sometimes light in weight. 



5. Light Seeds. 



greater rapidity of sprouting, or in higher total amounts, or in 

 both ; 3. The relative values of seeds of different colors vary with 

 each species, or sometimes with each sample. 



VIII. Influences of Latitude. 



Plants of high latitudes are more sensitive to heat and cold than 

 those of the same species growing nearer the equator ; i.e., they 

 start or vegetate relatively earlier in spring. This subject has been 

 investigated in several directions ; but, so far as the writer is 

 aware, it has not been pursued in this couutrj' in relation to ger- 

 mination of seeds. The following tests are incidental to this in- 

 vestigation, being a part of a general series of researches upon the 

 influence of latitude upon plants, but they are suggestive in this 

 connection. 



A sample of white dent corn was secured from the Alabama 

 Experiment Station, and samples of white and yellow dents were 

 obtained from the South Carolina Station. The germination of 

 these samples was compared with that of corn grown on the farm 

 of Cornell University. 



With corn from different latitudes, fifty kernels in each sample, 

 sown one inch deep in 12-inch seed-pans, sprouting was much the 

 most rapid in the New York corn, but differences in totals were 

 evidently not due to influence of latitude. The plants from New 

 York seed were by far the largest and most vigorous of any in the 

 test during the month which they remained in the house. The 

 Alabama seed gave the least vigorous plants, while South Carolina 

 seeds gave intermediate results. Fig. 7, from a photograph, 

 illustrates the New York and Alabama samples ten days after 

 sowing. 



Three other tests were made, with the same result. In one test 

 the sample from New York was represented by seed taken from a 

 crib of soft corn, yet this sample gave earliest results, though less 

 marked than in the other instances. A similar lesson appears to 



