Hill . — The Histology of the Sieve-Tubes of Angiosper ms . 279 
the threads are very numerous and are arranged in small groups in pits dis- 
tributed all over the wall, but at no time does there appear to be any callus 
associated with the threads. The cell-walls, however, are somewhat mucila- 
ginous and respond readily to swelling agents (Fig. 27, PI. XVII). 
In the case of the more typical Angiosperms the connexions between 
a sieve-tube and a companion cell are very different in character to those of 
the sieve-fields. In the latter, the connexions are fine slime-strings en- 
sheathed in callus-rods, and each half of the common cell-wall between the 
two tubes is of the same nature (Fig. 14, PI. XVII). But in the case of 
the wall between a companion cell and a sieve-tube the two portions of 
the dividing wall show a different composition. On the sieve-tube side of 
the lamella the wall is fairly broad, whilst on the companion-cell side it is 
narrow and appears to be somewhat different chemically from the walls 
of the sieve-tubes. The middle lamella thus comes to occupy an unsym- 
metrical or extra-median position (Fig. 60, PI. XVIII). 
The connexions between sieve-tubes and companion cells are in con- 
sequence also unsymmetrical in appearance, and moreover are found to be 
composed of two distinct portions. In the sieve-tube ‘ half’ of the wall the 
threads appear to be like those of the sieve-fields, being accompanied by 
callus, which reaches as far as the lamella ; whilst in the companion-cell 
‘half’ of the wall the threads are necessarily very short and are more like 
those in the walls of ordinary parenchyma cells, and there is no callus. The 
condition of affairs is thus analogous to that described for the Albuminous 
cells of Pinush where each individual ‘thread’ is similarly made up of 
two distinct portions. The character of the connexions between sieve- 
tubes and companion cells in Angiosperms is not nearly so easy of elucida- 
tion as is that of the similar connexions between the sieve-tubes and 
albuminous cells of the Gymnosperms, owing to the thinness of the walls 
in Angiosperms, and some difficulty exists as to the exact nature of the 
relation of the callus to the threads. During the spring and summer it is 
not possible in many cases to demonstrate callus at all in such a position, 
but in some plants a definite occurrence of callus in the sieve-tube portion 
of the common wall cannot be doubted (Fig. 29, PI. XVII, Figs. 51, 52, and 
60, PI. XVIII). 
Callus has been seen to extend as far as the middle lamella in winter 
material of Vitis (Fig. 52, PI. XVIII), but owing to the delicacy of the walls 
it is very difficult to make out its precise relation to the threads. It seems 
probable, however, that both in Vitis and Tilia^ and perhaps also in most of 
the Angiosperms, each whole ‘ connecting-thread ’ in the mature condition 
consists of a slime-string ensheathed in its callus-rod in the sieve-tube portion 
of the wall, which is then continued as, or rather is in contact with, an ordinary 
1 Cf. Hill, Ann. Bot., vol. xv, 1901, p. 600, Fig. 12, PI. XXXI; Fig. 12, PI. XXXII ; Fig. 23, 
PI. XXXIII. 
