Physiologie. 565 



The seleclive activity is independent of any living substance in 

 the seed coat. 



The coat of Xanthiinn can be removed and used as an osmotic 

 membrane, possessing a great advantagc over the coats investigated 

 by Brown and Schröder. 



The testa is composed of three layers, the outer of which can- 

 not function as a semipermeable membrane. The middle laj'er is 

 several cells thick, the inner layer one cell except in the chalazal 

 region. The last layer is probably the nucellar epidermis. 



B3' use of streng alkalies the inner membrane can be split 

 loose from the middle laj-er. Both la3^ers possess osmotic properties. 

 the inner layer in a higher degree than the middle one. Neither 

 layer is as efficient alone as when both are left together. The im- 

 pairing of the membranes maj'- be due to stretching. or to the ef- 

 fects of the macerating agent. 



The inner layer is nearly pure cellulose, unsuberized, but per- 

 haps containing some tannin. The middle coat contains more tan- 

 nin than the inner coat. 



The tannin does not form a continuous layer in the seed coat. 

 Moreover, treatment with tannin solvents does not destroy semi- 

 permeability. The evidence is adverse to Richard 's view that 

 semipermeability is due to tannin Compounds. 



Semipermeability has been demonstrated lor the seed coats of 

 a number of plants in six widely separated families. Man}^ dead 

 plant membranes may possess this property. 



The capillary and imbibition force of the embryo of Xanthiinn 

 as measured by the osmotic pressure of concentrated salt Solutions 

 is about 965 atmospheres when the seed is air-dry. 



An increase in the moisture of the embryo equal to 7 per cent 

 of its air-dry weight reduces the internal forces by 590 atmospheres. 



The unusual intake of water noticed with certain substances, 

 especially with certain acids and alkalies, is due largelj^ to the de- 

 velopment of osmotically active substances inside the semipermeable 

 membrane. 



There is some evidence unfavourable to Armstrong 's hydrone 

 the.ory of selective semipermeability. Jongmans. 



Theodoresco, E. C, Temperature mortelle pour quelques 

 diastases d'origine animale et vegetale. (Rev. gen. Bot. 

 XXVbis. p. 599—627. 1914.) 



II suit des resultats obtenues dans ce travail que les diastases 

 dessechees supportent un chauffage d'une demi heure ä des tem- 

 peratures assez elevees au-dessus de 100°; il 3' en a, tel le labfer- 

 ment, dont la temperature mortelle est superierre ä 191°. 



Puisque la veritable nature chimique des ferments nous est tres 

 peu connue, on ne peut pas donner une explication de ces faits. 

 Mais la plupart des diastases sont, comme tout porte ä le croire , 

 des substances colloidales, qui sont coagulees par la chaleur; par 

 consequent, il est permis de supposer que plus elles sont deshy- 

 dratees, plus leur coagulation a lieu ä une temperature ^levee; les 

 choses se passent comme pour les mati^res albuminoides. Cette 

 coagulation semble n'etre qu'un phenomene purement physique; 

 mais il n'est pas impossible, il est meme vraisemblable, que des 

 modifications chimiques plus ou moins profondes accompagnent le 

 phenomene physique. Jongmans. 



