Endosperm of Hordeum vulgare during Germination. 15 



"by such comparative experiments, to differentiate with certainty the 

 modifying action of micro-organisms from the autonomous action of 

 ihe endosperm-cells themselves. 



The action due to extraneous organisms always commences at the 

 surface of the "depleted layer," the cell-membranes of which this is 

 made up being softened, swollen, and ultimately disintegrated. This 

 -cytohydrolytic action then gradually extends to the membranes of 

 the amyliferous cells, and the proteid contents of the cells are also 

 involved in the change, which ultimately permeates the whole of the 

 endosperm. 



There is, however, a striking difference between the mode of pro- 

 gression of this bacterial action from that observed in " living " 

 degermed endosperms. In this latter case, as we have already 

 noted, the action is essentially centripetal, commencing under the 

 '' aleurone-layer " on the dorsal side, where this layer is intersected 

 by the " depleted layer," and extending peripherally and axially, but 

 more rapidly on the dorsal side. In the degermed "dead" endo- 

 sperms, on the other hand, there is no differential progression of this 

 kind, since the action, whilst progressing in an axial direction, does 

 not extend more rapidly along the peripheral than the central parts, 

 and does not show the slightest tendency to more rapid extension 011 

 the dorsal side, a tendency which is so strongly marked in "living" 

 degermed endosperms in water-culture, or in intact grains of barley 

 undergoing ordinary germination. It is only when the disintegra- 

 tion of the endosperm-contents under the action of micro-organisms 

 has proceeded to a very considerable extent that any notable amount 

 of erosion of the starch-granules is observable. This sometimes does 

 not occur for many days, a fact probably due to the bacteria not 

 secreting any special starch-dissolving enzyme as long as they are 

 well supplied with readily assimilable food material from other 

 sources. 



So far the conclusions are altogether opposed to the view that the 

 normal phenomena of endosperm solution and depletion, as they 

 occur in degermed endosperms in water-culture, can be explained by 

 the action of extraneous micro-organisms. It is true that, under 

 certain circumstances, the mixed growths of Bacteriacece which attach 

 themselves to the mutilated surface of the endosperm can induce 

 changes in the subjacent tissues by the projection into them of 

 certain enzymes, the products of their growth, but this action can, 

 with due care, be clearly differentiated from the normal action, 

 which is of quite a different character, and must be in some way self- 

 induced by the endosperm-cells themselves. 



Before considering how far the normal changes are dependent on 

 the vitality of any particular portion of the endosperm, we must 

 inquire if the phenomena are in any way due to enzymes pre-existent 



