662 Tho day.—On the Histological Relations between 
C. rejlexa parasitic upon Salvia , with corroborative evidence from C. europaea 
on Vitis. These hosts were found to be the most favourable for histological 
work; on the one hand their cell-walls are not so thin or their tissues so 
parenchymatous as, e. g., in Begonia or Impatiens , and on the other hand 
their cortex and pith are fairly free from the numerous crystals, fibres, stone 
cells, &c., which make the use of the freezing microtome so difficult with, 
e. g., the hop. In both cases, too, the compacted mass of dead cell-walls 
which at an early stage surrounds the haustorium is very soon absorbed ; 
in some hosts attacked by Cuscuta they are hardened, and form an almost 
insuperable barrier to histological investigation . 1 
The ingrowth of the haustorium is mainly due to the superficial cells 
covering its tip, which undergo a remarkable series of changes. They 
become greatly elongated , 2 and each forms a hypha-like element of very 
large dimensions. They have all the features characteristic of secretory 
cells: their walls are at first very thin, especially at the tip ; their proto¬ 
plasmic contents are very dense, and their nuclei are exceedingly large. In 
many cases the diameter of the nucleus of an invading hypha was found to 
be greater than the diameter of the whole cell invaded. 
Cell-division occurs at first chiefly in the cells immediately behind the 
hyphal cells , 3 but as the tips of the hyphae grow inwards their hinder 
portions become also subdivided, their elongated apices remaining, how¬ 
ever, unicellular throughout all the young and growing stages. As 
they extend further into the host plant it is common to find, in the 
tip of a single hypha, from two to five of the large characteristic nuclei 
(Figs. 46 and 47, PI. XLIX) ; these unicellular and multinucleate elements 
may branch at an early stage, but it is not until later that they become 
subdivided transversely and longitudinally almost up to the tip, each of the 
nuclei then being surrounded by a cell-wall. 
From each hypha a strand of cells is thus developed ; the further 
development of the different strands depends on their position with regard 
to the tissues of the host. The end of the invading haustorium in the 
cortex of the host may be compared with a brush of hairs, separated some¬ 
what from one another in the early stages. In C. rejlexa on Salvia and in 
C. europaea on Viiis the central hairs (hyphae) push straight through the 
vascular cylinder into the thick-walled pith of the host; they do not become 
subdivided for some distance from the tip, and generally remain small and 
undifferentiated. In a few cases, however, strands of sieve tubes were found 
to be developed from them even in C. rejlexa in the empty, innutritious 
1 This is still more strikingly the case in plants attacked by Lathraea squamaria , where the 
mass of dead cell-walls stains with lignin stains. 
2 Cf. Peirce, 1893, esp. Figs. 15, 16, 18, PI. XIV; Heinricher, 1894 ( Lathraea ), p. 325, Fig. 1, 
PI. IX, and Fig. 1, PI. X; Solms Laubach, 1868, Figs. 2, 5, 6, Taf. XXXV; Leclerc du Sablon, 
1894 (. Melampyrum ), Figs. 3, 5, &c., PI. I. 
3 See Peirce, p. 304, &c., and Fig. 5 g, PI. XIII. 
