ON THE STATISTICAL VIEW OF NATURE. 625 



tries to describe the. general mechanism under which, 

 the latter the individual steps and incidents ly which, 

 special events or phenomena proceed and are character- 

 ised. Earlier chapters of this narrative attempted to 

 give an account of the former, whilst the essentially 

 historical treatment belongs to another portion of the 

 work. The word history has generally been reserved for 

 records which deal with those events in which human 

 consciousness has played a large, if not an overwhelming, 

 part, and has been able to assist the observer by its own 

 accounts and representations. What should we know of 

 human life and human interests without them, and how 

 helpless in spite of minutest observation do we still 

 appear to be in understanding the life of the brute 

 and mute creation, even of the domestic animals, our 

 daily friends and companions ? But if history, as opposed 

 to statistics, really seems only possible where the living 

 voice or the surviving narrative of those who have de- 

 parted helps us to a true understanding of its incidents 

 and its meaning, it also imposes upon us the task of sift- 

 ing its value and trustworthiness critically. Mathe- 

 matics, logic, and statistics may do something to exclude 

 the actually impossible or the highly improbable from 

 a vast mass of material ; but more delicate criteria are 

 required in dealing with the accumulated testimony of 

 bygone ages. With an unerring instinct of what, in 

 addition to mathematical measurements, may be required 

 in order to accomplish this task, the nineteenth century 

 has not only nursed the scientific spirit and cultivated 

 its methods, but with equal diligence and originality 

 those other methods which lie at the foundation of 

 VOL. n. 2 R 



