394 TRANSACTIONS OF THE CANADIAN INSTITUTE. [VoL. VI. - 
The intermediate portion of the ciliated band is again cut longitudinally 
and the marginal portion obliquely. The arrangement of the ciliated cells 
in bands is clearly seen, the bands being much wider in comparison with 
the bands of intermediate tissue than in the last section, indications, in- 
deed, of their continuity being shown by the short but narrow and close- 
ly packed cells, which line the surfaces of each-intervening ridge.  Fi- 
nally, it may be stated, in a section still higher up one finds a perfect 
continuity of the bands of ciliated cells, the section cutting the marginal 
ciliated cells longitudinally. 
The general structure just described is essentially that described by 
previous authors and more especially by von Heider (1895). My inter- 
pretation of the various parts differs, however, somewhat from that given 
by von Heider. He recognizes the intermediate epithelium, but regards 
it as an endodermal layer separating the marginal ciliated ectoderm from 
the median lobe of the filament, which he identifies with the glandular 
streak of the Hexactinian filament. I shall return toa discussion of 
the nature of the epithelium of the median lobe later ; in the meantime 
I may point out what seems to me to bea fundamental error in von 
Heider’s interpretation. He regards the entire intermediate region of 
the wings as digestive in function, terming it the “ Driisenwulst” and 
identifying it with the endodermal areas of the glandular streaks which 
Willem (1893) has shown to be digestive. 
As a maiter of fact there are two very different kinds of epithelium in 
this intermediate region ; (1) that lining the furrows which run across it 
and (2) that occupying the intervals between the furrows. The former 
is exactly like the epithelium found at the free margins of the lamellz 
and is, indeed, continuous with this. In other words, the ciliated epi- 
thelium, which forms a continuous streak near the free edge of each 
lamella sends inward almost to the free edge of the mesentery a num- 
ber of prolongations which line the bottom of depressions on the sur- 
face of the lamella. Each of these prolongations is separated from its 
neighbours by a non-ciliated band (at least I have not been able to de- 
tect cilia in my preparations) of epithelium. It is this arrangement of 
the ciliated cells in transverse streaks which produces the characteristic 
transversely striated appearance of the lamellz noted by nearly all 
observers, and it is interesting to note that Thorell in 1857, with his 
usual accuracy, described the arrangement of the ciliated cells in trans- 
verse streaks in the ciliated bands of Metridium dianthus. 
The intermediate area cannot then be regarded as consisting wholly 
of endoderm since it is generally admitted that the ciliated cells are ecto- 
rn 
