Hill . — The Hisiology of the Sieve- Tubes of A ngio sperms. 249 
original substance which will become the pits of the young sieve-plate ; the 
meshwork thus tends to be of greater thickness than the rest of the plate. 
After this the substance occupying the meshes swells, forming projections on 
either side of the membrane, and stains with aniline blue, so that the contents 
of two tubes appear to be united by blue rods which occupy the axes of 
the meshes . 1 Gradually the whole of the substance filling the mesh takes 
on this blue coloration. Osmotic exchanges take place at first through 
the substance filling the meshes of the sieve-plate, but later communication 
between two adjoining tubes is effected by a definite mucilaginous filament 
traversing a canal or hole in the plate owing to the replacement of the 
original substance of the meshes by the invading filaments of the albuminous 
contents of the sieve-tubes. 
The parietal protoplasm also accompanies these filaments, so that there 
is communication not only of the slime but also of the protoplasm of the 
sieve-tubes. The exact way in which the change from the embryonic to 
the mature mode of communication takes place is by no means clear from 
Lecomte’s account. The development may not always go as far as this ; 
in Gymnosperms, for instance, and probably in many Angiosperms also, the 
meshes, according to his view, may always be occupied by the primitive 
substance of the membrane and may only afford places more suitable for 
osmotic exchange than other parts of the sieve-plate. 
Before criticizing these views of Lecomte on the development of the 
sieve-plate, it will be more convenient to give a brief resume of his opinions 
on the formation of the callus. He criticizes Russow’s results , 2 with which he 
entirely disagrees, and comes to the conclusion that the callus forming the 
callus-pads is due to the swelling of the thin layer of the substance which 
covers the cellulose network of the sieve-plate , 3 but that the cellulose net- 
work itself never gives rise to any callus. The first callus layer is an 
integral part of the membrane of the sieve-plate, and then this layer thickens 
at the expense of the albuminous contents of the tube. 
Thus this substance which fills the meshes of the young sieve-plate and 
also covers over the cellulose framework functions first of all — and in some 
cases always 4 — as the pathway for osmotic exchanges. The portions which 
are not transformed into mucous strings — i. e., the parts covering the 
cellulose network — then swell up to join the first layer of callus. 
According to Perrot’s account of the work of Lecomte and Leger , 5 the 
1 Perrot, ‘ Le Tissu cribli ', p. 38, and Fig. 19. 
2 Lecomte, 1. c., p. 263. 3 Ibid., p. 26 r. 
4 Ibid., p. 260. After the penetration of the meshes, the substance covering the cellulose 
network shrinks to an almost imperceptible layer, since communication is now direct, but in those 
Dicotyledons where he was unable to show the perforations he found the wall notably thicker than 
all the others. This is due to the fact that the osmotic changes take place exclusively here, and the 
non-cellulose portion retains the first stage of swelling seen in the young sieve-plates. 
5 Cf. Perrot, 1. c., p. 36 ; cf. Lecomte, pp. 248 and 260. It does not appear that Lecomte realized 
that the substance of his meshes which stains blue with aniline blue is really callus (v. Conclusion, 
S 
