188 
Thompson .— The Relationships of the . 
Discussion. 
The significance of these reticulate perforations in regard to the evolu¬ 
tion of the vessel depends on their relation to the scalariform type, and this 
in turn depends on the origin of that type. I pointed out in an earlier 
paper that the scalariform perforation may have originated in two ways: it 
may represent a slight modification of the primitive scalariform pitting, 
or may have been derived by the fusion in horizontal rows of circular pits. 
Miss Bliss has presented evidence that the scalariform pit and perfora¬ 
tion in angiosperms have originated in the second way—by the fusion 
of circular pits. It is necessary to assume, on this view, that the circular 
pits were originally formed from the scalariform ones by the reticulation 
of the latter, and that in angiosperms the reverse process has occurred, 
and the circular pits gone back again to scalariform by a fusion process. 
She believes also that the irregular nets, as seen in Rosaceae, &c., are 
simply a variation in this fusion process. 
Brown ( 2 ), on the other hand, holds that the scalariform condition 
is primitive in angiosperms, being a survival of the exactly similar pitting of 
the older gymnosperms and not a most remarkable instance of reversed 
evolution. The scalariform pit is being discarded for the circular one. It 
is first discarded on the side walls because, since these walls are in contact 
with a variety of cells differing in shape, size, &c., the scalariform pit loses 
its alinement. Under these conditions the circular pit is more adaptable 
and comes to prevail. On the end walls, however, which are in contact 
with walls like themselves, the scalariform type remains and becomes the 
scalariform perforation. 
Miss Bliss’s chief argument is the occurrence of intermediate stages 
between multiseriate and scalariform pits. Admittedly there are abundant 
examples of these transitions in angiosperms. But obviously the tran¬ 
sitions may be read in either direction. Bliss reads them from circular to 
scalariform and Brown from scalariform to circular. The decision must be 
made on other evidence. 
Brown states that in conservative regions the scalariform pit prevails, 
and that the multiseriate circular type characterizes the more specialized 
parts, while Bliss contends that the reverse is true. I am bound to state 
that in my experience with a large number of species Brown is correct. 
The order from the pith outwards as I find it is the familiar one of lower 
plants, spiral, scalariform, reticulate, multiseriate. And in a large number 
of species the scalariform pitting persists through the secondary wood with 
or without the circular type. It is, of course, possible in some forms to find 
scalariform pits in the older wood and compare them with carefully selected 
circular pits near the pith somewhere else. It is necessary to take into 
