222 POPULAR SCIENCE MONTHLY 



things. However, our knowledge of fowls always begins with the ex- 

 perience of certain sensations of sight with which sensations of hear- 

 ing and of touch may be associated. The sensations of sight, to which 

 we will limit ourselves for the present, are by no means completely 

 identical. According to its distance from us, we see the fowl as large 

 or small; with changes in its position and its movements its contour 

 is very different. As, however, we observe that these differences pass 

 continuously into one another without exceeding certain limits, we 

 overlook them and confine ourselves to certain other peculiarities (legs, 

 wings, eyes, beak, comb, etc.) which remain constant and do not change. 

 The constant properties we gather together as a ' thing ' : the changing 

 ones we call the states of this thing. Among the changing ones we 

 distinguish those which are dependent upon ourselves (e. g., distance) 

 from those upon which we ourselves have no immediate influence (e. g., 

 position and movements). The former we call the subjectively vari- 

 able part of our experience, whereas we term the latter the objective 

 variability of the thing. 



To ignore the subjectively and objectively variable portion of our 

 experiences, while we retain their constant parts, and to combine the 

 latter into a single unity, is one of the most important operations which 

 we base on our experiences. We term this procedure abstraction, and 

 its product, the constant unit, a concept. Obviously this procedure 

 contains arbitrary as well as essential parts. Quite arbitrary, or rather 

 accidental, is the fact that according to the state of our attention, our 

 training, nay even our whole intellectual make-up, quite different parts 

 of any given experience reach our consciousness. We may overlook 

 constant components and notice changing ones. But all components 

 become of necessity objective as soon as we have noticed them. After 

 once seeing the fowl black, it is no longer in our power to see it red. 

 It follows that in general our knowledge of corresponding character- 

 istics is less extensive than it might be, inasmuch as we have never 

 noticed all that correspond. Our concept is, therefore, poorer at any 

 given moment in components than it might be. To search for these 

 hitherto overlooked components of a concept and to prove them a con- 

 stant part of the corresponding experience is one of the never-ceasing 

 labors of science. 



The other possibility, viz., that certain components which do not 

 prove to be constant have been incorporated into a concept, also occurs 

 and leads to another problem. These questionable components may, on 

 the one hand, be eliminated from the concept if further experiences show 

 that the remaining ones are contained in them ; or, on the other hand, a 

 new concept may be formed by including the constant components and 

 eliminating the inconstant ones. Thus for a long time the white color 

 was a part of the concept ' Swan.' When the black swans of Aus- 



