Palcontoloi^ical Speculations. — Grafacap. 95 
new ones equally small appear. It is true that there are alter- 
ations and waves of advance and retreat, that a group of 
strongly developed organisms may be followed by less de- 
veloped examples of the same groups. But looked at broadly 
and carrying the eye forward rapidly the impression is unmis- 
takable that there is a tendency to molar expansion, skeletal 
growth, and that from some maximum point there may be 
recession and that both seem essentially independent of en- 
vironment, and are inexplicable on the assumption of the "Sur- 
vival of the Fittest." 
To begin with the linguloid and obelloid shells can be 
traced upward to Trimerella through a series of increasing 
size and changing internal hard parts or shell buttresses, while' 
nowhere do large sized individuals signalize the first appear- 
ance of a family. Dr. Clarke in the first particular instance 
offers convincing evidence. He points out (Pal. Vol. VIII., 
Geol. Surv., N. Y.) that a line of succession from a linguloid 
shell through two divergent lines can be traced to the large 
heavy Trimerellas with internal platform, prominent cardinal 
areas, solid and lengthened umbo. One line is through Lin- 
gulella, Lingula, Lingulops, Lingulasma, the second through 
Oholella, Oholiis, Elkania Dinobolus. Whether this is a true 
genealogical series or not it is a very clear proof of assumption 
of size and skeletal mass through successive formations, of a 
group of ecardines, though Lingulops is a small shell, and 
serves the purpose only of advancing internal structure. Lin- 
gulasma, on the other hand, is a quadrate strong and well de- 
veloped shell, and, as Dr. Clarke shows, is "to be regarded, in 
unison with other features, as evidence of progress in the 
assumption of the characters of the large thick-shelled Trim- 
erellids." In the other hypothetical succession Oholella is a 
small sub-circular shell with an area; Obolus is a large shell, 
Elkania is a doubtful link and a small shell, Dinobolus is a 
large platform-bearing shell and in the American Niagara (D. 
Conradi) presents approximative features to Trimerella. 
Again Lingulepis is a primordial linguloid shell, and is lim- 
ited in its distribution to Cambrian beds, but increase of size 
is symptomatic even within these limits. The L. niinima and 
L. antiqua of the N. Y. Potsdam are small, the former an in- 
conspicuous shell, whereas L. pinniformis, the western form 
