98 THE AMERICAN’ MIDLAND NATURALIST 
bearing eight cells on each side, counts of three were frequent. 
Usually this group of cells is separated from the frontal epi- 
thelium by a single gland cell which was usually very large, 
2 microns wide by five microns long. ‘The mucous globule 
full of granules was always found extruded from the top of 
this cell probably due to fixing. The cells of the latero-frontal 
epithelium were columnar in shape but were larger than those 
of the frontal epithelium. Their average length was four 
microns, while the width varied from one to two microns 
being greater in those cells nearest the distal end. This vari- 
ation in the thickness of these cells is probably due to less 
crowding of these nearest the inter-filamentar junction since 
the proximal cells are on the curve of the filament (see fig. 3). 
The cytoplasm of these cells is not so granular as that of the 
frontal epithelium. The nucleus has the same ellipsoid shape 
as in the frontal epithelium but is slightly larger and located 
nearer the center of the cell. The neuclolus is always very 
distinct in these cells. 
The epithelial cells from this point gradually assume a more 
spherical shape and although bearing cilia, the cilia are short, 
usually less than one micron in length. About one-half the 
distance from the edge of the filament to the inter-filamentar 
junction the cells have assumed a spherical shape and the nu- 
cleus is also practically spherical at this point. From this 
point on the cells gradually lengthen in the other plane giving 
rise to the pavement epithelium lining the ostia and water 
tubes. The nucleus also gradually lengthen until the ellipsoid 
form appears again but this time the long axis of the ellipse 
is parallel to the base instead of perpendicular as before. 
These cells are from three to four microns long and an aver- 
age two microns in width. The ratio ef length to width is 
therefore three to two, or rarely four to two, and never be- 
comes as great as_that in the frontal or latero-frontal epithe- 
hum. It should be remembered however that this ratio is the 
converse of that mentioned first since the cells are now flat- 
tened in the horizontal direction instead of the vertical, that 
is the former width has by gradations now become the length 
of the cell and vice versa. Not all of these pavement cells 
retain their cilia but the number is so variable that no stand- 
