258 Hill — The Histology of the Sieve-Tubes of A ngiosper ms. 
show a broad band of phloem with sieve-tubes in all stages of development, 
from their earliest formation to their disused and collapsed condition. 
In material of this kind the sieve-tubes are larger and have thicker 
walls than those of the young shoots; moreover, the developmental changes 
appear to proceed more slowly, and are more easily observed in tissues 
gouged out of the main stems. 
The preservation of the material is probably the most important matter 
in connexion with this kind of histological research, and the sieve-tubes of 
Angiosperms appear to offer much greater difficulties than those of the 
Gymnosperms. This seems to be due mainly to the character of the sieve- 
plate, which, owing to the size of its pores, allows a shrinkage of the 
sieve-tube contents to take place on cutting into a stem, and produces the 
characteristically contracted appearance which is nearly always seen in 
fixed sieve-tubes. The same kind of difficulty is experienced with developing 
sieve-tubes, since their watery contents are enclosed in only a thin proto- 
plasmic sac ; on fixation plasmolysis readily takes place, and the contraction 
of the protoplasm is liable to distort the true structure of the sieve-plate 
with its developing slime-strings. 
The artificial appearance of the fixed sieve-tube was pointed out by 
Fischer, 1 but his method of fixation by boiling tends to produce effects 
which do not represent faithfully the living structure. 
Certain experiments, which have been carried out on living plants of 
Cucurbita with the intention of fixing the sieve- tubes in an uninjured con- 
dition, with the usual reagents, were to a certain extent successful, for they 
show a more or less uniform distribution of the slimy contents within the proto- 
plasmic sac, and in this respect show some agreement with Fischer’s results. 
Methods. 
The methods employed in this investigation were based on those 
published by Gardiner 2 in 1898. They were on the whole similar to those 
employed in the examination of Pinus , but certain modifications in detail 
have been introduced as occasion has required. 
The material was killed either in very small pieces, consisting of cambium 
and phloem only, or in the form of sections cut from fresh frozen material. 
All killing agents and fixatives were in aqueous solution, and comprised 
solutions of Iodine in Iodides, Picric acid, and mixtures of such solutions 
with other reagents. Solutions containing corrosive sublimate and Chromic, 
Acetic or Osmic acids were not found useful as killing agents. 
It has been found necessary, as in the demonstration of ordinary 
connecting-threads, to induce a certain amount of swelling of the killed 
tissues in order to obtain satisfactory staining reactions, and for this reason 
1 Cf. P . 24s. 
2 Gardiner, Proc. Camb. Phil. Soc., 1S98, p. 508. 
