46 DR H. SCOTT-LAUDER ON SOME 
authority on eryptogamic botany, and he kindly assisted me 
to name and classify the species. The result, with figures, 
may be found in the 7ransactions of that Society for 1864, 
vol 12. 
The general characters of the family Chextocerex are as 
follows (figs. 2 and 10):—They are filamentous, the frus- 
tules are united to each other by the interlacing of tubular 
awns which spring from the angles or circumference of the 
contiguous cells, which in front view are quadrangular, and 
in side view oval (Chextoceros) (fig. 3) or circular (Lacteri- 
astrum), fig. 11. The breadth of frustule is from 3/99” to 
300. 
The awns are usually many times longer than the breadth 
of the frustule, in section either cylindrical, quadrangular, or 
hexagonal, and having bead-like tubercles or short spines 
arranged in a spiral manner around the length of the awn, 
or with a cellular structure in some species. 
In the perfect filament the awns of the terminal valves, 
that is to say, the two valves of the original frustule (fig. 1) 
produced from the sporangium or auxospore, are usually 
shorter and stouter than the intermediate ones, and often 
differ from them in shape (fig. 2). 
The process of multiplication of the frustules in a filament 
is similar to that common to all diatoms. The endochrome 
shrinks away from the sides of the cell, forming a spherical 
central mass with strongly defined outline. This and its 
nucleus divide into two equal portions, the connecting hoop 
of the frustule at the same time increasing in breadth until 
the frustule is double its original length. The two masses of 
endochrome each secrete a new valve, a line of demarcation 
is formed in the hoop, which apparently contains very little 
silex, new awns sprout at the line of demarcation, and thus 
two frustules are produced, each consisting of a new and an 
old valve. 
At certain seasons, or under certain unknown con- 
ditions, this process takes a different direction. The con- 
densed endochrome, instead of dividing, secretes a silicious 
envelope, in shape quite unlike the parent frustule, forming 
an endocyst within the frustule, probably of a sporangial 
nature (fig. 4). The term endocyst has been suggested by 
