EEPOET OX THE TOXIC ATA. 
79 
is visible, and the bladder cells haA*e become polygonal by mutual pressure (PL V. fig. 17). 
The second form of cell is found vhereA’er a small patch of matrix can be seen. They are 
the test cells, and are small (O'OOS to O'Ol mm.), ellipsoidal, fusiform, or stellate in shape, 
and consist of a large nucleus surrounded by a little protoplasm which sometimes 
stretches out to form long delicate processes. These cells stain deeply with carmine. 
Under a low power (50 diameters) this test tissue looks like a fine network minutely 
dotted OA’er (PL Y. fig. 14). The meshes are formed by the bladder cells, of which several 
layers may be in focus at once, and the dots represent the nuclei of both kinds of cells. 
When magnified 200 diameters the structure is apparent (PL Y. fig. 15). The outlines 
of the large bladder cells seem to cut each other constantly ; this is due to the fact that 
more than one layer is visible. If care be taken to focus three or four cells lying in the 
same j)Une, it will be seen that they neA'^er intersect or open into one another, and that 
except when they are A’ery croAvded a thin layer of matrix lies betAveen them. In some 
places under this power the tissue looks as if composed of large protoplasts (the patches 
of hyaline matrix, containing the small fusiform cells with large nuclei) united to one 
another by radiating processes and leaA'ing large lacunse between (the bladder cells). 
This appearance is very striking in some preparations stained with picrocarmine 
(PL Y. fig. 17). 
A higher power (350 diameters) shows better the nuclei of the bladder cells and their 
parietal position. Coarse granules are generally Ausible in these nuclei. Throughout the 
rest of the head the inAmsting mass has essentially the above described structure ; the 
bladder cells, however, are generally not quite so large, and are much more crowded, 
being polygonal in shape ; sometimes they form regular hexagons, showing no matrix 
except here and there at the angles. The central part of the head, as mentioned above,, 
is quite sjDongy in appearance, being channelled out by the Avascular appendages. Under 
a low power (50 diameters) a section of this region presents a curious appearance 
(PL Y. fig. 16). The large sj)aces (about 0'2 mm. across) reduce the tissue to an irregular 
reticulum, the thick bars of which liaA^e the ordinary structure of bladder cells, &c.„ 
Avhile here and there are seen threads too narrow to contain a bladder cell ; these ar& 
formed of the matrix, Avith its small spindle-shaped test cells. 
Towards the base of the colony, on account of the greater number of vascular 
appendages which have to be accommodated, the spongy area is greater in extent and 
the amount of tissue in the spongy part is less, being reduced to bars of nearly uniform 
thickness which enclose spaces of hexagonal form, thus making a regular netAAmrk 
(PL YIII. fig. 11). 
Over the whole head the surface layer of the test tissue differs slightly in structure 
from the central part (PL Y. fig. 17). It contains no bladder cells, but is formed merely 
of the homogeneous matrix and the small fusiform or stellate test cells. This layer can 
be stripped off as a delicate membrane from the surface of. the tissue beneath. The 
