Hill . — -The Histology of the Sieve-Tubes of A ngio sperms . 283 
growth and obliteration of the threads 1 and also as to the secondary changes 
which they may undergo. It is not easy to understand how Strasburgers 
plasmodesmic pseudopodia 2 could be formed coincidently on either side 
of a separating membrane with the regularity with which the connecting- 
threads are found in the two halves of the common wall. 
Summary. 
1. The young cell-wall, which will develop into the sieve-plate, is at 
first a homogeneous, pitted membrane. 
2. The pit-closing membranes of the young sieve-plate are crossed 
either by small groups of fine protoplasmic threads, e. g. Wistaria chinensis 
and Cucurbita pepo , or in some cases apparently by only a single thread, 
e. g. Vitis vinifera. 
3. Callus makes its appearance in the form of little basins lining the 
pits of the developing sieve-plate. It appears to be due to the alteration of 
the superficial layers of the cellulose membrane at these points by ferment 
action. Eventually this change extends over the surface of the whole 
membrane. 
4. With the inception of the callus-change the fine thread or threads 
of the young sieve-plate, as the case may be, begin to be bored out to form 
slime-strings apparently by a ferment, which at the same time affects the 
pit-closing membrane in the immediate vicinity of the threads and converts 
it into callus. 
5. The enlargement of the slime-strings thus formed continues, with 
the final result that a single large slime-string is found occupying the place 
of each pit of the young sieve-plate. In the one case the single string is 
due to the enlargement and fusion of a group of strings, in the other to the 
enlargement of a single string. 
6. The slime-string is in all cases enclosed in a protoplasmic tube, 
which passes through the callus-lined pore of the sieve-plate. 
7. Subsequent formation of callus appears to be due to protoplasmic 
activity and not to cellulose changes. The callus is laid down over the 
sieve-plate by the protoplasm, apparently in definite layers, after the manner 
of cellulose, and at length so constricts the pores of the sieve-plate that 
translocation is prevented. 
8. The sieve-pores in a large callus-mass are not obliterated, though 
much attenuated. In the plants examined, whose sieve-tubes function for 
1 Gardiner, Proc. Roy. Soc., 1900, p. 187. Also Hill, Phil. Trans. Roy. Soc., cf. PI. XXXI, 
Figs. 1 and 7 ; PI. XXXIII, Figs. 21 and 23. Gardiner, Proc. Camb. Phil. Soc., 1907, p. 209. 
2 Strasburger, 1 . c., p. 503, considers that this meeting of the plasmodesma is no more strange 
than the formation of corresponding pits. But the formation of the corresponding pits is surely due 
to the fact that there are already groups of threads at these spots, and that it is necessary for the 
greater efficiency of the threads that the pit-closing membrane should be thin. 
U 2 
