LOGIC OF RECEPTS. 63 



with all his might. Evidently the previously straight line 

 must yield somewhat in the middle, whatever the weight of 

 the fly, who was, in fact, thereby brought into position F', to 

 the right of the first one and a little 

 higher. Beyond this point, it might 

 seem, he could not be lifted ; but the 

 guy being left fast at ^, the spider 

 now went to an intermediate point c 

 directly over his victim's new posi- 

 tion, and thus spun a new vertical 

 line from c, which was made fast at 

 the bend at (C, after which a d was 

 cast off, so that the fly now hung 

 vertically below c, as before below a, but a little higher. 



" The same operation was repeated again and again, a new 

 guy being occasionally spun, but the spider never descending 

 more than about halfway down the cord, whose elasticity 

 was in no way involved in the process. All was done with 

 surprising rapidity. I watched it for some five minutes 

 (during which the fly was lifted perhaps si.x inches), and then 

 was called away." 



Without further burdening the argument with illustrative 

 proof, it must now be evident that the " ore " out of which 

 concepts are formed is highly metalliferous : it is not merely 

 a dull earth which bears no resemblance to the shining sub- 

 stance smelted from it in the furnace of Language ; it is 

 already sparkling to such an extent that we may well feel 

 there is no need of analysis to show it charged with that sub- 

 stance in its pure form — that what we see in the ore is the 

 same kind of material as we take from the melting-pot, and 

 differs from it only in the degree of its agglomeration. Never- 

 theless, I will not yet assume that such is the case. Before 

 wo can be perfectly sure that two things which seem to the 

 eye of common sense so similar are really the .same, we must 

 submit them to a scientific analysis. Even though it be 

 certain that the one is extracted from the other, there still 



