Cuscuta and its Host. 
657 
more clearly understood. The sieve plates could still be located, and the 
callus could still be stained ; it was thus possible to obtain reciprocal pre¬ 
parations of the same element, in the first place with protoplasmic con¬ 
nexions and slime strings well shown, in the second with the walls stained 
(see Figs. 76-8, PL LI). 
I. Notes on the Histology of Salvia (sp. ?) (perennial). 
(a) The sieve tubes of Salvia are found to resemble closely those of 
Vitis 1 in the general outlines of their development. The small size of the 
narrow terminal sieve plates makes them difficult to investigate, but it has 
been possible to make out sufficient of their structure for the present purpose, 
and it is interesting to find so close a resemblance between the sieve tubes 
of the Vine and other large plants heretofore examined, and those of a small 
plant like Salvia . 
Fig. i, PL XL IX, represents a young terminal sieve plate in section ; it is 
seen to be unpitted and traversed by a number of fine protoplasmic threads, 
on some of which a minute swelling is visible at the middle lamella. On 
the upper surface of the plate faint specks of callus have already appeared 
at the ends of the threads. The extension of the callus change is seen in 
the terminal sieve plate in Fig. 2, where each thread is enclosed in a callus 
rod. Figs. 3, 4, 5, represent other sieve plates in the same and later stages. 2 
Surface views at these stages showed each string to be in the centre of 
a hole traversing the plate and surrounded by a ring of callus. Shallow 
pits are usually formed in the young sieve plates, but are not universally 
present even in the quite mature plate. When present, each pit Is always 
occupied by a single thread or slime string, the method of development 
being that of Hill’s second type. 3 
All the older sieve tubes in this material were in the stage shown in 
Fig. 6, where a large mass of callus is seen on either side of the plate. 
Lateral sieve plates were rare ; they appeared to resemble the terminal 
sieve plates in all particulars (Figs. 4 and 8). Fig. 7 represents a lateral 
plate in surface view; the ends of the callus rods are arranged in groups, 
and the groups are distributed over the plate, collected together in well- 
defined areas. In one lateral sieve plate (Fig. 8) median nodes were clearly 
seen separating the two halves of the callus rod; these were also present 
in the terminal sieve plates, but are not shown in any of the figures. 
The sieve fields are rather more abundant in the tangential than in the 
radial walls. 4 In surface view numerous elliptical depressions are seen in 
the wall; each depression is a shallow pit, the membrane of which is 
traversed by numerous fine threads (Figs. 10 and 11). Each thread is 
bored out and transformed into a slime string enclosed in a separate callus 
1 Hill, 1908. 2 See description of plates. 
3 Hill, 1908, p. 274, Text-figs, io, n, 12. 4 Cf. ibid., p. 268. 
