162 The American GcologlM. September, 1892- 
correspond in<i- zones of the several series. By this niethod the 
law was estaljlished thai the composition of a fossil fauna changes 
on passing geographically from one place to another. Upon 
tracing single species across these sections it was leai'ned that the 
mutation of the species, not only may be recognized on passing 
vertically upward through a continuous section, but that the more 
direct line of succession was often deflected laterally so that the 
immediate successor of a particular fauna of one section was 
found not directly above it in the same section, but at a higher 
horizon in a section ten or twenty miles distant. This shifting of 
faniKis was taken as actual evidence of migration and was inter- 
preted as the result of oscillations of level. 
The examination of the remarkable fauna of High Point, at the 
southern end of Canandaigua lake (the locality of which was first 
shown me by Mr. J. M. Clarke), furnished me with still further 
clue to the solution of the origin of the faunas. I recognized, at 
once, upon seeing it, that it was related to the Iowa Devonian, and 
diflfered widely from the typical upper Devonian of New York, in 
the midst of which it lay. 
Further anal^'sis of the fauna led to the discovery that the 
species peculiar to it apparently had their ancestors in the middle 
Devonian of Europe rather than in any middle Devonian of Amer- 
ica. With this stage of progress I examined the fauna peculiar 
to the Tully limestone. Much confusion had been thrown about 
it by the publication of a large number of species as ••known'" 
Tully fossils. Special search was made in original localities with 
the result of eliminating a large number of reported species which 
were found immediately below the true Tully limestone in the cal- 
cai'eous termination of the Hamilton, where the typical Hamilton 
fauna is very abundant, and the true fauna which 1 described as 
the Cuhoides fauna, was carefully compared with that of every 
locality in the world of which I could find report of its presence. 
The result showed that in eastern America where the TuU}' ap- 
pears, the fauna of the Cuboides zone begins abruptly, and from 
it upward, all through the upper Devonian, is a fauna closely re- 
lated in its species with the upper Devonian of Europe, Russia, 
Siberia, China antl British America, and down as far as Iowa in 
the interior, the Nevada Devonian also showing close affinities 
with this type or fauna. 
