HILGARD—NATURAL SERIES. 183 
never was the offspring of synthetic argument, but of compar- 
ison ; and although authors have adopted their arrangement, 
yet, in contradiction to their very principles of total corres- 
the Orders are subdivided on partial criteria, often 
gratuitously selected and quite inefficient. 
The pues, by which I arrived at my results, was one of 
Iyg0 um, Datisca, and the like), after the fundamental rule 
atural affinity, sepia to setiglinssions viz., closest total 
snes of characters; the dignity of each character being 
partly as an impression of judgment derived from their com- 
parative values of co constancy and variability on the individuals 
and within the species, and withal from the relative part they 
viz., 4 f 
assumed, a posteriori, subse t apositions of 
species, genera, orders, sho on the fundamental leading prin- 
ae gay an 
cen, $e one ey conch abate 
