190 Thompson .— The Relationships of the 
production, and consequently the various reticulate conditions cannot repre¬ 
sent a variation of such a fusion process. Reticulate perforations of the type 
found in Compositae, Ericaceae, &c., appear to be modified reticulate pitting 
which is commonly associated with scalariform, and which in some cases is 
derived from the latter as a passing phase in the production of multiseriate 
pitting. 
The uniform nets of Cordia , Tecoma , &c., are relatively very rare and 
sporadic, and occur in highly specialized families (Borraginaceae, Bignonia- 
ceae, Verbenaceae), so that they appear to have no significance in the 
evolution of the vessel as a whole. Moreover, there is no evidence of 
pit fusions in these cases. The individual pores are of the same size or 
only slightly larger than bordered pits. The whole net appears to drop 
out at once. They are found in only a few of the vessels. It seems probable 
that they represent a special condition. Apparently the vessel-segments 
start to form alternate pitting on all walls, and then the membranes drop 
out at the ends. In some cases the net seems to be due to the fact that the 
end of one vessel-segment abuts on the side wall of another overlapping seg¬ 
ment. The end becomes perforated, but the side on the other element 
retains its alternate pitting, in many cases losing the membranes. Some¬ 
times two adjoining vessels have lateral communications, and it is not 
unnatural that in such conditions the reticulate condition derived from 
alternating pits should occur. 
As for the fusions in Cydonia , Potentilla , &c., in some cases at least 
they appear to be due to the loss of reticulations in the same way as 
the bars disappear from scalariform perforations. It is to be noted that 
frequently both scalariform and reticulate conditions are found in the same 
perforated area which is breaking down to form the simple perforation. 
The scalariform pit has been transformed into the multiseriate through the 
reticulate condition, and there appears no reason why this transformation 
could not have begun in occasional forms before the pits became perforate. 
At any rate the reticulate condition frequently accompanies the scalariform, 
and may break down to produce the simple perforation in the same way that 
the scalariform does. But even if we grant that they are real fusions 
of multiseriate bordered pits, it seems that they could have little signifi¬ 
cance. They are of such rare and sporadic occurrence, and are found 
in such specialized forms (chiefly herbs), that their existence can be of little 
real importance, in view of the evidence that the scalariform type is primitive 
and the widespread occurrence of transitions between scalariform and simple 
perforations. 
It is undoubtedly true that in the great majority of angiosperms the 
uniperforate vessel has been derived from the scalariform by the loss of 
the bars, and the balance of the evidence shows that the scalariform per¬ 
forations are the result of a slight modification of the primitive scalariform 
