696 LAURA FLORENCE 
not in the anterior, but toward the posterior, half of the stomach, in the 
region of the bend cephalad; and in the last numbers of the series it was 
found throughout the whole slender part of the stomach. 
In the cells containing the greatest number of granules, some were 
seen resting on the basement membrane and a few appeared to be lying 
among the muscle fibers outside the membrane, but sufficient evidence 
to prove that they had passed through the membrane unchanged is wanting. 
Whatever may have been the fate of these granules, they disappeared 
from the cells, leaving in their places numerous vacuoles among which 
the first traces of secretion were seen. The secretion accumulated in 
the form of a compact mass, resembling a ball of thread, whose surface 
layer takes a deep stain while the axis remains almost clear. This ball 
pressed against the free ental border of the cell, pushing it into the lumen 
and finally rupturing it. 
The above experiments show that absorption and secretion are carried 
on by the same cells. In every section some cells, evidently in a resting 
stage, are seen, but it is not clear whether the cells pass through this 
stage after each secretion or at longer intervals. The formation of the 
secretion appears to begin at the close of absorption, and, as the study 
of the starved lice suggested, its excretion is stimulated by hunger, so 
that it is already present in the stomach when. the blood is ingested. 
No attempt has been made to investigate the exact nature of the granules 
or the changes they may undergo, as this would necessitate a long series of 
experiments with various reagents, such as were carried out first by 
Fischer (1899) and later by Murlin (1902). 
If lice be fed as in the previous experiment, and blood smears be made 
from the stomach contents at intervals of one hour and stained with 
Wright’s stain, the gradual action of the epithelial secretion on the blood 
can be followed. Within one hour after feeding, the red cells become 
vacuolated and fat globules appear, but the leucocytes and the platelets 
are evidently not affected. The changes in the red cells continue until 
only an amorphous mass remains, which, in sections stained with hema- 
toxylin and eosin, can be recognized as a mass of brownish granules. 
No blood platelets have been seen in any but one-hour smears. At two 
hours the nuclei of the leucocytes are intact but their cytoplasm has been 
attacked; there is a gradual change in its staining reaction, and after three 
hours it takes the basic stain, appears light blue, and can be distinguished 
