from the Bochy Mountains. 255 



fragmentary condition — to belong to Permian types, and 

 declares the evidence to be decisive on this point. 



The animal remains consist almost exclusively of insects, 

 and are two thirds as abundant in species as the plants — an 

 exceptionally large ratio in beds where both occur. These 

 insects form an assemblage wholly different from anything 

 before known, and, in contradiction to what Mr. Lesquereux 

 says of the plants, clearly belong to types of a more modern 

 character than any the Palaeozoic series has yet disclosed. It 

 is not often that one may speak so positively in the discussion 

 of fossil insects, especially when not a single one of the 

 species and only the smaller portion of the genera found have 

 been previously known. But in this case all but two or 

 three of the specimens obtained (some eighty in inimber) 

 belong to a group which of all Pala30zoic insects has re- 

 ceived the most attention, namely the cockroaches. This 

 great preponderance of cockroaches, and the fact that the 

 few known genera found in this collection have hitherto 

 been discovered only in Carboniferous and Permian rocks, 

 would lead us at first to refer the beds in which they occur 

 to one of the Paleozoic series ,• but the presence of the other 

 forms, and even the characteristics of those Avhich are refer- 

 able to Carboniferous and Permian genera, unmistakably 

 point to a later horizon. 



Palceozoic cockroaches are distinguished from living types 

 by the complete interdependence of two of the veins of the 

 front wing, and by the fact that the anal veins of the same 

 wings invariably impinge upon the inner margin, and never, 

 as in existing forms^ upon the anal furrow. For these 

 ancient types the name of Palseoblattari^e has been proposed, 

 and all PalEeozoic cockroaches whose front wings are pre- 

 served (and we know them almost exclusively from these 

 organs) fall into this group. k:)0 far as I can discover there 

 is not a single exception to this difference between ancient 

 and modern types. Since this was first stated five years ao-o 

 the number of Paleeozoic species has been increased 25 per 

 cent., and it is still true. 



Inthe paper in which these points were first discussed no 

 allusion was made to Mesozoic cockroaches, as none had 

 been found in this country, and the illustrations we possess 

 of the European species are in many cases by no means 

 sufficient to expose their structure ; their study was therefore 

 left until the imperfection could be remedied. It was, how- 

 ever, recognized, thougli not stated, that Pala3oblattariie exist 

 in Jurassic rocks; it is showMi, for instance, by figures of 

 Wealden species on the fifth page of Brodie's work ' On tlie 



