lo NATURAL APPEARANCES 



of the vegetable turf, is a thing eafily to be conceived, Uke v^hat 

 happens in thofe ftripes the fecond year, vv^hen I have feen an 

 abundant crop of a certain fpecies of mufhrooms in the track. 

 Had animals of a particular fpecies been found there, in the exa- 

 mination of the foil in thofe v^ithered tracks, a rafh conclulion 

 might have been formed, in erroneoully attributing as a caufe 

 for the appearance, what was truly an effedl or confequence of 

 the thing in queftion. 



It is always making a flep towards the difcovering the caufe of 

 a phenomenon, when caufes which, with fome degree of pro- 

 bability, have been afcribed to an event, are found to be uncon- 

 nedied with, or to have no affinity to it ; for this is the natural 

 method of inveftigation, by examining the affinities or rela- 

 tions of things, and rejedling thofe as properly related, where 

 there is found a difcrepancy. Thus, as there is no effedl with- 

 out its proper caufe ; fo, in proportion as a greater number of 

 events are found to be unconnedled with an appearance, fome 

 kind of approach is made towards that by which the natural 

 appearance is to be explained ; but in cafes where events are 

 multiplied or numberlefs, every approach of this kind is only 

 negative j and fuch a method of inveftigation, while it may be 

 the means of difcovering the thing in queftion, only fliows 

 that what we want is not attained. This, however, if made 

 with full convi(flion, is no contemptible ftep in natural philo- 

 fophy, where, next to the inveftigation of the proper order in 

 events, it is of the higheft importance to avoid, or to corre(5V, 

 the improper conneiflion of them. 



The explanation of the phenomena, in the prefent piece of 

 natural hiftory, either by thunder or the operation of infeds, 

 without having obferved the a(5lual connecflion of thofe diffe- 

 rent events, is merely conjedlural, as would be equally the re- 

 fuftng to admit for explanation a known caufe, which, though 

 not adually obferved as conne(5led with the event in queftion, 



had, 



