18 STUDIES IN ROOT-PARASITISM. 
command, they do not appear to resemble the double cystoliths 
described for Cansjera by van Tieghem.* 
They occur in isolated cells and the manner of deposition 
appears to be rather that of ordinary calcium oxalate crystals 
than of normal cystoliths. They form masses of clearly 
defined tabular crystals of calcium carbonate or a mulberry-like 
mass with blunt and frequently branching projections (Plate 
VI, figs. 2and 3). Inone case a quite different form is assumed. 
In the outer cortex of some sections the crystals of calcium 
carbonate have taken on the appearance of spherocrystals 
(Plate VI, fig. 4). Such crystal masses, instead of belonging 
to one cell, project into two or three from their common angle. 
At the place of junction a substance is found which stains deep- 
ly with methylene blue. In the haustoria where these occur, 
the radiating crystals were found in the outer cortex, while 
the crystal masses in the inner cortex had the normal form 
described above.+ 
In a few cases the walls of the cells in which the crystal 
masses occur have become sclerotic (Plate VI, fig. 5). These 
thick-walled cells call to mind the sclerotic cystoliths found in 
Champereia.; 
Isolated thick-walled cells are found scattered throughout 
the cortex of Cansjera haustoria. Their usual position is 
* Van Tieghem, Recherches sur la Structure et les Affinités des Thymeleacées et des 
Peneacées. Ann. a. Se. Nat. Ser. VII, Bot., T. XVII, 1893: pp. 249-257, Cansjera, Cham. 
pereia et la Famille des Opiliacées, 
+ The facts concerning these supposed radiating crystals of calcium carbonate are as 
follows. They were found in sections of several haustoria attacking Capparis horrida roots. 
In these haustoria the ordinary crystal masses were present in abundance in the usual places 
and the radiating crystals were confined to the ends of the cortical folds. Uponthe addition 
of dilute acetic acid all the crystals disappeared rapidly, with the evolution of bubbles of gas, 
but it was not possible to assert that any of this gas came from the masses of radiating crystals, 
because it made its appe trance at points of the sections far removed from any crystals. 
The radiating crystals do not exhibit the true structure of sphzrocrystals as described by 
Hansen. They are simply groups of crystals projecting into several cells from a basis of amor- 
phous substance which was readily stained. There was no outer shell or skin visible. It 
is possible that they do not occur in fresh material, but have been slowly formed, like sphzro- 
crystals, in the alcoholic material, (See Ad. Hansen, Ueber Spherocrystalle, Arb, Bot, Inst, 
Wiirzburg, III, 1888.) 
t Solereder, Syst, Anat. d, Dicotyledonen, 1899, 
