Endosperm of Hordeum vulgare during Germination. 9 



commencing with extremely dilute solutions, which were gradually 

 increased in strength until their influence on the germinative power 

 of the seed was just perceptible. The germicidal effect of such a 

 solution was then tested on degermed grains in water-culture. At 

 one time extremely dilute solutions of formaldehyde and of acid 

 potassium fluoride offered some hope of success in this direction, but 

 neither of these substances on further investigation gave a sufficient 

 differential action to be of any practical use. 



In the experiments of 1890 (Brown and Morris, loc. cit.) the dis- 

 turbing effect of micro-organisms was minimised by restricting the 

 time of the experiment as far as possible, and by sterilising the 

 culture-media, and we have seen that Hansteen relied on killing the 

 adherent germs with a solution of copper sulphate, and on the 

 employment of strict antiseptic methods, even to the extent of carry- 

 ing out all the operations in a germ-free atmosphere. 



We have made experiments in order to see how far such a treat- 

 ment with copper sulphate effects sterilisation of the integuments, 

 the grain after such treatment being incubated in contact with vege- 

 table infusions. The results have clearly shown us that although 

 such a procedure may retard the subsequent development of Bac- 

 teriacece and moulds, it is impossible by means of it to ensure a com- 

 plete destruction of all the germs adherent to the palese, unless the 

 treatment is sufficiently prolonged to destroy, or at any rate to mate- 

 rially reduce, the germinative power of the embryo. 



Since any process which will affect the .vitality of the embryo 

 cannot be without some similar influence on the endosperm, there is 

 thus introduced an element of uncertainty into all subsequent pro- 

 cesses w r hich maybe devised for determining whether the amyliferous 

 cells are living or dead. 



Extreme refinements for avoiding air-sown organisms are obviously 

 of little efficacy when complete initial sterilisation of the exterior of 

 the grain cannot be ensured. Nevertheless, many of our experiments 

 have been carried on with precautions of this kind, but have not 

 yielded better results than those made in covered dishes with sterili- 

 sation of the culture-media and apparatus. 



In all experiments with endospermous seeds deprived of their 

 embryos both Hansteen and Pfeffer have, very properly, laid great 

 stress on the necessity for providing for a rapid removal of any 

 possible products of change in the isolated endosperm as fast as they 

 are formed, but these observers have apparently entirely overlooked 

 the fact that this was fully insisted upon and provided for in the 

 earlier experiments described by one of us in 1890.* The plan 

 adopted was to insert the proximal ends of the degermed grains 

 into small holes made in a thin mica plate, which was then floated 

 * ' Chem. Soc. Journ.,' 1890, Trans., p. -i8l. 



