138 
Journal of Agricultural Research 
Vol. XXIII, No. 3 
has been altered in composition, possibly because of the local spreading 
of the enzym that made the pore. 
The formation of the minute opening in the two walls places the two 
osmotic membranes of host and parasite in direct contact, and the sub¬ 
stance of the haustorium mother cell immediately begins to push through 
into the host cell (PI. 2, E). As has been mentioned, the osmotic pressure 
in the fungus is probably higher than in its host, and this entrance may 
take place automatically by the extension of the fungous membrane as 
water passes into the fungus from the weaker solution in the host cell. 
This extension of the fungous cell into the host is accompanied by a 
corresponding invagination of the host protoplast. The appearance 
suggests strongly that the plasma membrane of the host is not broken 
but merely cupped by the invader. 
A slightly larger haustorium is seen in Plate 2, F. The peglike hau¬ 
storium of the earlier stage has now become differentiated into its two 
parts, the body, a dense globular mass showing no details of its contents, 
and the slender neck. The haustorium mother-cell is still dense in 
contents, and, as in Plate 2, E, there are concentrated darker staining 
masses within it. 
Plate 2, G, is typical of a slightly later step in the process. The 
penultimate cell has its pair of full-grown nuclei and is pushing out 
branches just below the septum. Much of the contents of the terminal 
cell have passed into the haustorium, and in the less dense remainder 
we discover that the two small nuclei are still present. The young 
haustorium consists of a dense ball of fungous cytoplasm which stains a 
uniform deep red and a narrow neck joining this ball to the parent cell. 
The haustorium is still invested with host cytoplasm and still gives the 
impression of having stretched the peripheral layer of the cytoplasm of 
the host inwards and pushed it ahead into the central vacuole, forming 
a sort of pocket. If this impression is correct, the young haustorium is 
in one sense still outside of the living protoplast of the host, having 
merely indented its plasma membrane. The fact that the turgor of the 
host cell is not destroyed, but is, on the contrary, slightly increased, 
would also suggest an unbroken osmotic membrane. 
Plate 2, A, c and d, shows a somewhat more advanced stage. The cell 
at d giving rise to the haustorium is now empty except for a rim of cyto¬ 
plasm concentrated at its distal end near the point of exit into the 
haustorium. The haustorium c is larger and the two nuclei, or at least 
the nucleoles, are clearly visible in it. It is also beginning to expand, as 
is seen from the small vacuole within it. The rest of the haustorium is 
homogeneous in appearance. 
Still later (PI. 2, H) the parent cell appears to be empty, and the 
haustorium now contains the complete protoplast of a fungous cell. 
The haustorium has undergone a rapid expansion by the absorption of 
water from the host and now presents an appearance similar to that of 
the cell from which it came. The cytoplasm has much the same structure 
as before except that it is more open. One of the two nuclei is clearly 
visible in it. 
Plate 2, I, shows an older U-shaped haustorium seen from the upper 
end of the U. The cytoplasmic investment is exceptionally heavy and 
includes a chloroplast at a and a small accumulation of cytoplasm about 
the neck. Looking into the ends of this haustorium, one sees that the 
peripheral portion is much more loose and open in structure than the 
center. Even the older, full-grown haustoria often have a denser core. 
