Inexact Observations. 3 



pretended constancy or variability of species, declare 

 with imposing certitude that, ia laying down the 

 boundaries between species, represented by them as 

 systematic units, they were supported by results de- 

 rived from the cultivation of plants under observation. 

 Nay, even those who were in a position to proceed 

 experimentally, did not hesitate, in their eagerness to 

 support a preconceived idea, to issue as true coin 

 fictitious observations. If to writers such as these we 

 add those others who had no intention to deceive, but 

 who rather deceived themselves in consequence of 

 improper methods of research or carelessness of obser- 

 vation, then indeed must all confidence in the state- 

 ments of past writers fail us ; and this the more, as we 

 have no possible means of correctly separating the 

 chaff from the wheat. 



Doubtless it is a very convenient plan to select, by 

 your own standard, out of the chaos of existing state- 

 ments, such as can serve to support a given hypothesis, 

 and to adopt these as "recognised facts;" whilst you 

 represent those which cannot be brought into harmony 

 with it as the results of erroneous experimentation and 

 of ignorant or inexact observation. But in the build- 

 ing up of a science no good can come from such an 

 arbitrary choice of materials. Considering, therefore, 

 the utter uncertainty we are in as to the use of the 

 older statements, there remains no other course open 

 to us than to begin the whole work afresh, and so 



