1947] 
Early Insect Life 
67 
state that these fossils are very fragmentary, the largest 
specimen being only about a millimeter long. Many parts 
of the head and thorax have been found, but since only 
the first three abdominal segments are known, there is no 
evidence that the collembolan “spring” was present, or 
that the abdomen consisted of but six segments, as in 
living Collembola. Consequently, although such eminent 
authorities as G. H. Carpenter and If. Womersley have 
accepted the specimens as Collembola, I believe that doubt 
about their relationship will exist until additional fossils 
have been found showing those two characteristic fea- 
tures of the Collembola. As a matter of fact, as Imms 
and others have pointed out, the Collembola are not true 
insects — that is, although hexapodous arthropods, they 
arose independently from pre-insectan types. The possi- 
ble occurrence of Collembola in Devonian rocks, there- 
fore, has no direct bearing on the earliest record of true 
insects. 
The other discovery of an insect in rocks older than 
the Upper Carboniferous was announced by the French 
paleontologist, Dr. Pierre Pruvost, in 1919. The speci- 
men concerned was found in Lower Carboniferous strata 
of Nova Scotia (Horton’s Bluff), and identified by Pru- 
vost as a member of the extinct Order Palseodictyoptera. 2 
However, examination of this fossil, which is now in the 
Peabody Museum at Yale University, convinces me that 
it is in reality part of a plant stem. A similar conviction 
was expressed by Dr. David White, 3 then paleobotanist of 
the United States Geological Survey, who studied the 
specimen shortly after it was found, although he did not 
publish his conclusion. 
According to our present knowledge, therefore, the 
earliest unquestionable insects occur in Upper Carbon- 
iferous strata. This first occurrence is not extensive, but 
it does indicate the approximate time of origin of the 
insects. Let us return for a moment to the geologic time- 
table, and examine in more detail the Upper Carbonifer- 
ous Period (Figure 2). This represents an interval of 
2 La Faune continentale du terrain Houiller dn Nord de la France, p. 283. 
3 In a letter dated May 11, 1914, and on file in the Peabody Museum. 
