IMPERFECTION OF THE GEOLOGICAL RECORD 45 



are disposed to look with scepticism on the information 

 obtained by this laborious but certain process, and to suppose 

 that they are being presented with imaginary "restorations." 

 I think it right here to copy a remark of a German botanist, 

 who has felt himself called to criticise my work : " Dawson's 

 description of the genus {Psilophyton) rests chiefly on the 

 impression made on him in his repeated researches," etc. 

 " He puts us off with an account of the general idea which he 

 has drawn from the study of them." This is the remark of a 

 closet naturalist, with reference to the kind of work above 

 referred to, which, of course, cannot be represented in its 

 entirety in figures or hand specimens.^ 



As to the precursors of the Carboniferous flora, in default 

 of information already acquired, I proceeded to question the 

 Erian or Devonian rocks of Canada, in which Sir William 

 Logan had already found remains of plants which had not, 

 however, been studied or described. Laboriously coasting 

 along the cliffs of Gaspe and the Baie des Chaleurs, digging 

 into the sandstones of Eastern Maine, and studying the plants 

 collected by the New York Survey, I began to find that there 

 was a rich Devonian flora, and that, like that of the Carboni- 

 ferous, it presented different stages from the base to the summit 

 of the formation. But here a great advance was made in a 

 somewhat unexpected way. My then young friends, the late 

 Prof. Hartt and Mr. Matthew, of St. John, had found a few 

 remains of plants in the Devonian, or at least pre-Carboniferous 

 beds of St. John, which were placed in my hands for descrip- 

 tion. They were so novel and curious that inquiry was stimu- 

 lated, and these gentlemen, with some friends of similar tastes, 

 explored the shales exposed in the reefs near St. John, and 

 when they found the more productive beds, broke them up by 



^ Solms-Laubach, *' Fossil Botany." A pretentious book, which should 

 not have been translated into English without thorough revision and 

 correction. 



