90 



SCIENCE. 



[Vol. XIV. No. 340 



(Vaughan) — were, that (l) the quantity of vvater'applied to seeds 

 under test may make a remarkable difference in the results ; (2) that 

 sprouting is decidedly more rapid in tests which receive less than the 

 usual amounts of water given in greenhouses (this is markedly the 

 case in all the tests, with the exception of three indifferent and 

 comparatively unimportant instances) ; (3) that the per cent of 

 sprouting is much greater, as a rule, in the dryer tests ; (4) that 

 the addition of water above the amount to keep the earth simply 

 moist is injurious ; and (5) that the wide differences between the 

 results of the wet and moist tests are not necessarily due to the 

 rotting of the seeds in the wet tests (this is shown in the two tests 

 with cucumber-seeds in which the dryer tests gave similar or even 

 smaller totals than the wet tests). 



In the tests with carrot, sprouting was remarkably more rapid 

 in the dryer pan, and the per cent of sprouting was also very much 

 greater, amounting to 47 per cent. Fig. 3. from a photograph, rep- 

 resents this test at its conclusion. 



With the Lima bean, the per cent of sprouting was over 70 per 

 cent greater in the dryer pan. This was due to the fact that more 

 of the beans rotted in the wet pan. On May 22, twenty-six of the 

 beans sown May 4 were rotten in the wet pan. Only six were 

 rotten in the dryer pan, and ten were sprouting. It is known that 

 seeds with a slight surface abrasion often germinate better than 

 those which are uninjured ; but this test indicates that great care 

 must be exercised to water such seeds sparingly, as they are more 



soaked seeds are sown earlier than the dry ones. If this advance 

 in period of sowing is added to the date of sowing of the dry seeds 

 it will be found that dry seeds as a rule sprout fully as early as 

 soaked seeds, and many times much earlier. 



3. The total amount of sprouting does not appear to be influ- 

 enced by soaking. 



4. Similar results are not to be expected from all species of 

 plants. 



IV. Influences of Character of Soil. 

 It is well known that texture of soil often has much to do with the 

 germination of seeds in the field. Soils which bake, which become 

 very dry, or which hold too much moisture, always tend to give a 

 poor " stand " of crop. But the soils used in houses are such as 

 to occasion no thought of their influence upon germination ; yet 

 there are cases in which such soils cause variation in seed-tests. 

 This was particula'rly marked in a lot of beans tested this spring. 

 Samples happened to be sown at the same time in potting soil on 

 a bench, and under a cloth on the surface sand. Those in soil 

 gave much poorer germinations than the others. Other sowings 

 were therefore made from the same lot at given depths in sand for 

 purposes of comparison. The figures cannot be presented in the 

 limited space of this article, but it was found that sproutings were 

 in some cases nearly twice as many in sand as in potting soil. 

 More beans rotted in the soil than in the sand. The soil had not 

 been sifted, and it contained some manure ; yet it was only four 



likely to rot. Fig. 4, from a photograph, represents this test on 

 May 20. 



III. Influences of the Soaking of Seeds before Sowing. 



It is a common practice in both field operations and seed-testing 

 to soak seeds in water before sowing. Several tests made indi- 

 cate very clearly the leading results of this custom. In this con- 

 nection it is interesting to study results with the Geneva seed- 

 tester, which tests seeds by soaking them. A number of tests 

 were made with the Geneva tester in comparison with sowing in 

 potting soil in forcing-house. The results, which are too extended 

 to be detailed here, indicate that higher sprouting tests are given 

 by the Geneva tester than by planting under known conditions. 

 Ten tests in each case with Marblehead Mammoth cabbage-seeds 

 gave an average germination of 88 per cent in the tester, against 

 77.6 per cent in the soil. Tlie earliness at which the sprouting is 

 visible in the tester renders testing expeditious ; but it must be 

 remembered that full germination cannot often be secured in the 

 apparatus. (Cf. § IX.) 



The conclusions drawn from the tests — two with carrots (Early 

 Forcing and Vermont Butter), four with tomatoes (one Green 

 Gage, three Belle), one with turnips (Early Six Weeks), two with 

 radishes (Early Scarlet Globe), and one with onions (Giant Yellow 

 Globe Rocca) — were as follows : — 



1. Great gain in rapidity of sprouting, counting from the time of 

 planting, may be expected as a rule, if seeds are previously soaked 

 in water; and the longer the seeds are soaked, within reasonable 

 limits, the greater is usually the gain in rapidity of sprouting. 

 This fact is interesting, in face of the experience that very profuse 

 watering after sowing gives an opposite result. (Cf. § II.) 



2. This gain in rapidity of sprouting in soaked samples is really 

 fictitious, however, inasmuch as germination actually begins in the 

 soaked seeds before the dry samples are sown. In truth, the 



inches deep on the bench, and it would seem that the drainage 

 was good. Tests in this direction warrant the following conclu- 

 sions : I. Variations in results of testing may sometimes be ex- 

 pected in consequence of character of soil in which the tests are 

 made ; 2. In the present instance, low results in potting soil, as 

 compared with tests in saitd, appear to be due to the greater 

 amount of water held in the earth, causing the seeds to rot. The 

 results may therefore be studied in connection with those upon the 

 influence of varying amounts of watering. (Cf. § II.) 

 V. Influences of Light. 



Darwin, in his " Cross and Self Fertilization " (American edi- 

 tion), p. 13, says, " On other occasions, from the want of time, the 

 seeds, instead of being allowed to germinate on damp sand, were 

 sown on the opposite sides of pots, and the fully grown plants 

 measured. But this plan is less accurate, as the seeds sometimes 

 germinate more quickly on one side than on the other. It was, 

 however, necessary to act in this manner with some few species, 

 as certain kinds of seeds will not germinate well when exposed to 

 the light. . . . This occurred in the plainest manner with the seeds 

 of Papaver vagicm and Delphinium cotisolida, and less plainly 

 with those of Adonis astivalis and Ononis minutissima. Rarely 

 more than one or two of the seeds of these four species germinated 

 on the bare sand, though left there for some weeks ; but when 

 these same seeds were placed on earth in pots, and covered with 

 a thin layer of sand, they germinated immediately in large num- 

 bers." 



Of late years there has been more or less said concerning the 

 sowing of seeds for test upon the surface of soil, and covering 

 with glass in order that every seed may be watched ; and certain 

 seed-testing apparatus have been devised upon this principle. It 

 appears from Darwin's experience that with some seeds grave 

 errors may occur from this practice, and further evidence of the 



