- RAYMOND: NOTES ON THE ONTOGENY OF PARADOXIDES. B2Y 
forms must therefore belong to another species, which is here de- 
scribed as P. haywardi. The largest cranidium of the rimmed form 
obtained is 35 mm. long, while the smallest cranidium with the brim 
and no rim is 19 mm. long (M. C. Z., No. 22, Pl., fig. 3). On this 
specimen the part of the brim in front of the glabella is 2.25 mm. wide, 
or nearly 12% of the total length. Ona specimen 34 mm. long it is 4.5 
mm. long, or 13%, while on a large cranidium, 103 mm. long, it is 
12 mm. wide, or 11%. On the numerous cranidia between the smallest 
and largest it varies from 10% to 138% of the length, showing that 
while it grows wider during the growth of the individual, it is relatively 
about the same width in all cranidia above 19 mm. long. What it 
might be in smaller specimens we have as yet no means of knowing. 
The significance of this wide brim on the cranidium of P. harlani is 
best appreciated after studying the ontogeny of Paradoxides. 
ONTOGENY OF PARADOXIDES. 
The smallest specimen of Paradoxides known is that described by 
Barrande as Hydrocephalus saturnoides (Systeme Silurien du centre 
de la Bohéme, 1852, 1, p. 380, pl. 49). This specimen is slightly over 
1 mm. long, the cephalon is 
oval, and makes up five sixths 
of the totallength. The glabella 
is large, oval, makes up most 
of the cephalon, extends to the 
front of the head, and has no 
glabella furrows, though there 
is a median longitudinal furrow. | | 
The palpebral lobes form the Fig. 1.— Hydrocephalus saturnoides, show- 
lateral margins of the cephalon, ing two of the stages of development. 
but judging from the appearance —gtabella Compare with Plate, fz. 9. 
of the cranidium, the free cheeks 
would have been present even at this early stage, had the specimens 
been complete. The occipital segment is prominently set off from 
the rest of the cephalon, and extends to the long, intergénal spines 
which cross it at right angles. One thoracic segment and a pygidium 
are present. [rom the inferred presence of the free cheeks and the 
presence of a thoracic segment, it is evident that this is not a protaspis, 
but that several moults have already taken place. 
The second specimen described by Barrande is 1.33 mm. long, the 
