210 TRANSACTIONS OF THE [may 18, 
of development: 1. The early larval stage—non-agnostiform— 
when it possessed about 1-3 somites ; 2. The later larval stage 
—agnostiform—when it had about 4-5 somites; 3. The adult 
condition, when it had about 6—T somites, and the three main 
lobes of the rachis were developed with 4—5 somites in the pos- 
terior of the three. 
The above remarks are based wholly upon the three types of 
Agnosti that appeared in Europe and America simultaneously 
in the first sub-fauna of Paradoxides. Tous the special features 
of the other two sections of the genus are of less moment, as 
they appear to be simply a progressive effacement of those that 
are found in the earlier types; and it is in accordance with this 
that they appeared later in time, the Parvifrontes in the third 
sub-fauna of Paradoxides, the Leevigati scarcely before the fourth, 
while these latter culminated in the sixth, seeing that the beds 
are called the zone of Agnostus levigatus. 
In these later types of the Paradoxides beds both the glabella 
and the rachis are effaced, so that scarcely any means remain of 
distinguishing the heads from the tails in these smooth tests. 
Nevertheless it is not such aberrant types that carry on the 
genus Agnostus to a later time, but modifications, more direct, 
of the primitive types; for though the Regii disappeared from 
the stage of life in the early Cambrian time, Fallaces and Longi- 
frontes survived in the Ordovician seas, and the former lived on 
to its close. Thus this type which was the most abundant when 
the Agnosti appeared in the seas of Sweden, Wales and eastern 
Canada was the last to disappear when the genus was finally ex- 
tinguished at the close of Ordovician time. 
Tullberg, in his classical memoir on the genus Agnostus,* 
divides it into four groups, distinguished by well marked and 
easily recognized characters.+ The arrangement is such as to 
greatly aid in determining the species of this large genus, so 
characteristic of all parts “of the Cambrian from the ‘Olenellus 
zone upward. ‘The greatest development of the genus is in the 
Paradoxides zone, and especially the upper part. 
The following are Tullberg’s sections : 
Longifrontes. ‘ Distinguished by a manifest extension of the 
glabella and rachis, which commonly is rather long. Test some- 
times smooth, sometimes the cheeks are furrowed, sometimes the 
test both on the cheeks and pygidium is studded with elevated 
points. The we fold is cenerally narrow. The cheeks in 
*Om Mionuaians arterna i de Cambriska aflagringarne vid Andrarum. S. A. Tullberg, 
Stockholm, 1880. 
+It may also with propriety be divided into five sections as the Limbati contains 
two very distinct types. 
