NOT KNOWLEDGE. 335 



may be a little different, and yet we may observe 

 no difference in the result, let us scrutinize it as 

 we may. But that is owing 1 to the limit of our 

 observation being always within the limit of na- 

 ture, so that when the difference of the circum- 

 stances (all of them being known) eludes our ob- 

 servation, so does the difference of the result. 



Upon difficult subjects it is astonishing to what 

 an extent the multiplication of these little differ- 

 ences will in the end mislead us, if we do not keep 

 the whole chain carefully in view ; for in twenty 

 successive occurrences we may attend carefully to 

 each, and compare it with the one immediately 

 before it, and find the very same apparent simi- 

 larity in each of those comparisons ; and yet the 

 differences, unseen in the individual cases, may so 

 mount up in the aggregate, as that the last may 

 be unlike the first or even the very reverse of it. 

 With careless observers who are satisfied with a 

 few of the external and more obvious circumstances, 

 that is much more frequently the case ; and as 

 they who publish their opinions or conclusions to 

 the world, are not always the most close and ac- 

 curate observers, that is the reason why so many 

 errors have crept into the science and the systems 

 of Natural History ; and as many of these errors 

 are fortified by high authority, and all of them by 

 some authority (for there always are people so 

 forward in their belief that the very fact of being 

 in print is an authority to them), they are very 

 difficult to be reduced. 



The only means of doing that is, by going back 

 to the beginning of the series, that is, to nature 

 itself; and hence the superiority of knowledge 

 which we get from our own actual observation, to 



