of the Endosperm during Germination. 453 



closing membranes, (the " aggregated threads " of Kohl) which are 

 concerned with the translocation of the cellulose-dissolving enzymes 

 of the seed. In support of this view he brings forward the case of 

 the endosperm of Phoenix reclinata, which, he says, possesses pit 

 threads only ; and here he finds that enzymes do not travel by 

 means of the threads, but attack the thick cell walls which are 

 devoid of any threads and leave the pit-closing membranes un- 

 altered. 



But an examination of the endosperm of species of Phoenix 1 

 by our methods proved as Pfurtscheller discovered that both pit 

 and wall threads occur 2 , though the latter are usually found only 

 in special parts of the endosperm, viz. in the walls of cells near 

 the periphery of the seed and of those immediately surrounding 

 the embryo. 



In the course of germination the thick walls undoubtedly 

 appear to be attacked by the ferment and dissolved independently 

 of the threads, but further examination seems desirable to deter- 

 mine whether the initial stage of the conduction of the ferment is 

 or is not similar to that of Tamus, where it would appear that the 

 contagion of ferment action having been once communicated by 

 the threads, the subsequent rapid disorganisation and solution 

 extends over the whole area of the wall independently of them. 



The pit threads are still visible after enzyme action has pro- 

 ceeded for some time, but we do not find that the pit-closing 

 membrane remains unaltered, since it swells considerably and 

 finally becomes disorganised like the rest of the wall. 



A similar state of things was noticed in the endosperm of 

 Galium, where well-marked threads can be seen in the cell walls 

 around the cotyledonary cleft, though the walls have been so 

 altered by the enzyme that they no longer give the reactions of 

 the normal unaffected walls of the seed. 



The manner of disorganization of the cell walls of the endosperm 

 of Phoenix, which seems to have so little relation to the presence of 

 threads, appears to us to offer additional proof that the connecting 

 threads are not concerned primarily with the translocation of 

 enzymes, for those portions of the thick cell walls, which do not 

 possess " connecting threads," can be dissolved by ferment action 

 in the absence of pores by which it might be conveyed into the 

 substance of the walls. Where, however, threads are abundantly 

 distributed throughout the walls they do no doubt afford convenient 

 channels along which the ferment may pass, and by means of 

 which it may reach the older layers of the cell wall, but it seems 



1 The species used were P. sylvestris and P. dactylifera. 



2 Pfurtscheller, N. Jahrb. KK. Franz-Jos. -Gymn., Wien, 1883, p. 63. Kohl, 

 loc. cit., p. 367, mentions that "solitary threads" occur in the endosperm of 

 Phoenix, 



32—2 



