THE UNFATHOMED UNIVERSE 13 



scriptivc and did not admit of formulation in terms of ab- 

 stract concepts, as astronomy, for instance, does. Yet there 

 is in natural history a kind of description which is just 

 as essential in its own place as is mechanical or dynamical 

 description namely, historical description. We would not 

 buy a horse on the strength of a description in terms of the 

 dynamics of particles, partly because we could not go far 

 in the way of checking its accuracy, but mainly because of 

 our shrewd conviction that the essential thing is to know 

 what we can about the horse's history. Similarly, our 

 science of the horse must include not only its whole architec- 

 ture from skull to blood crystals, not only the consensus of 

 its active parts from brain to phagocytes, but also its char- 

 acter and its individual and racial becoming. Even within 

 the sciences of the inorganic, when dealing, for instance, 

 with the geological interpretation of scenery or the establish- 

 ment of the solar system, the description must be genetic 

 or historical. It is an interesting point that, just about the 

 time when Physics began to proclaim emphatically that its 

 office was to describe not to explain, Natural History in 

 Darwin's hands passed emphatically from description to 

 historical explanation. 



4:. Limitations of Natural Knowledge. 



Science makes so many permanent discoveries, which are 

 never contradicted though often transcended, that she ac- 

 quires an assured confidence which has only been equalled 

 by that of Theology. For this very reason it is useful that 

 she should be ever examining herself. One of the famous 

 balance-sheets was that made by Du Bois-Reymond in his 

 lectures on the Seven Riddles of the Universe and on the 

 Limits of Natural Knowledge. He confessed that the 



