NOTES AND ABSTRACTS. 



183 



tartaric acids are excluded, as well as dry alcohol, ether, chloroform 

 and acetone when the seedcoat is dry. Oxygen also does not diffuse 

 through absolutely dry seedcoats. On the other hand, when the 

 seedcoat is wet, alcohol, ether, iodine, potash, and sodium hydrate, 

 various nitrates and acids (nitric, acetic, lactic, and citric) enter slowly 

 or rapidly." 



This relative influence on absorption is exercised by the inner 

 membranous layer of the seedcoat (probably originally the nucellar 

 membrane), which is of nearly pure cellulose, but perhaps contains 

 some tannin. 



Treatment with tannin solvents does not destroy semipermeability. 

 Six different families of plants — Alismaceae, Grasses (Barley, Wheat, 

 and Oat), Sugarbeet, Rosaceae (Peach, Apple), Leguminosae, and 

 Compositae — also possess semipermeable seedcoats. 



Even minute injury to the seedcoat (not perceptible to the naked 

 eye) may result in allowing free entrance to substances which would 

 otherwise be excluded. 



These results have an important practical bearing on experiments 

 concerning germination, especially in the use of stimulants to germina- 

 tion and of fungicides and insecticides, for many observers have not 

 known that the seedcoat is semipermeable. 



The author also points out that the capillary and imbibition 

 force of the embryo of Xanthium amounts to about 965 atmospheres 

 when the seed is air-dry, and 590 atmospheres with an increase of 

 7 per cent, in the air-dry weight of the embryo. 



These high values are compared with the results of Rodewald, 

 who found that dry starch on swelling develops a pressure of 2523 

 atmospheres. — G. F. S. E. 



Seedweight of Bean, Correlations regarding. By J. Arthur 

 Harris [Beih. Bot. Cent. xxxi. Abt. i, Heft i, pp. 1-12 ; 4 pi. and 

 4 figs.). — ^This is a quantitative study of the factors influencing the 

 weight of the bean seed so far as intraovarial correlations are concerned. 



Five series, comprising 23,312 individually weighed seeds, are 

 shown in condensed tables, and the results are examined biometrically. 



The number of ovules is practically negligible as a factor influencing 

 seedweight (coefficient from — -0071 to + -0052 grams). Weight of 

 seeds decreases as the number of seeds per pod increases. Correla- 

 tions average only — '096. For relative number of seeds as compared 

 with ovules per pod the correlation is — •073. The chances of an 

 ovule becoming a seed are greater the nearer it is to the stigmatic 

 end, especially in small pods. Seedweight also increases towards the 

 stigma, but this is not always the case. — G. F. S. E. 



Silver-Leaf Disease. By F. T. Brooks, M.A. (Jour. Agr. Sci. 

 V. pp. 288-308; June 1913 ; 2 plates). — The author records a large 

 number of experiments with silver-leaf. The author regards the 

 silvery appearance of diseased foliage as a symptom possibly produced 



