6 Early Trilobites of the Cambrian (Recks — Matthew. 
decidedly from the preceding genus, and show that the two 
cannot be considered as belonging to one family. Tullberg's 
division of the Agnosti into four sections is eminently conven- 
ient and useful; both as showing the natural relations of the 
species, and also for determining the position and relations of 
new species. It may not be a perfectly natural one, as there 
is at least one species at Saint John which by its furrowed and 
punctate checks and thin test is related to the Longifrontes^ 
but does not possess a furrow dividing the cheeks in front of 
the glabella; it therefore is not a Longifrons according to Tull- 
berg's definition. So also the sections of the Llmbati^ which 
Tullberg designates respectively "Regii" and "Fallaces," in the 
light of the development of the trilobites from their earlier 
stages seem worthy of more prominent place than Tullberg 
has given them; in the above table the species that belong to 
these sections are shown, as well as those that fall into the 
"Longif routes" and "Parvif routes." 
Microdiscus though first named by Dr. Emmons is properly 
based on the species (J/, ptmctatus) described by J. W. Salter 
from the Welsh Cambrian measures; this species was found 
to have four joints in the thorax, but Mr. S. W. Ford has since 
found that a species of the Georgian fauna (J/, speciosus) has 
three joints; this number again is further reduced by the dis- 
covery of the thorax of J/. Daxvsoni which proves to have 
only two joints; at least this was the number found in an ex- 
ample three quarters grown. A thorax with 3-4 joints should 
therefore be ascribed to Microdiscus. 
Microdiscus shows an advance of development above Agnos- 
tus in many respects, and cannot be regarded as of the same 
family. The pleurae are built to give facility for sliding one 
over the other at their extremities, and the pygidium has the 
power of increasing the number of the rings of its rhachis. This 
feature is best seen in the latest of the three forms of the species 
noticed In the above table which when one quarter grown pos- 
seses seven rings. This is the number in the adult of the earlier 
species J/. Dawsoui and also according to Mr. Saltt-r marks 
M. pHHctatiis the type of the genus, to which M. pulchellus 
is closely allied, but at maturity our species has ten or eleven 
rings in the rhachis of the pygidium. Three rings is the normal 
