92 Rev. J. Wills on Mr. Stewart's Explanation of 



not be the result of any real perception, but is easily explicable by the process 

 already described ; the form, hue, and motion of the flame had been so associated, 

 that the incident wanting in the representation was supplied, before the judgment 

 could come into operation.* Of this nature are those cases also, already slightly 

 adverted to, of faces framed by the imagination out of accidental lines. Let me 

 dwell a moment on this, for it is one of a large and diffusive class of results, to 

 all of which the same explanation will apply. I mean that class of expressions 

 and effects which must in part be referred to the fancy of the observer. The 

 expression of the human countenance offers an instance in which several varied 

 qualities of human character seem combined with certain dispositions of form, in 

 such a manner, that while the expression is instantly presented to the observer, he 

 can in few instances, and then but partially, and by much nicety of observation, as- 

 certain the precise arrangements of feature to which the characteristic expression 

 is due. I shall not encumber the case by an analysis of the origin of such combi- 

 nations ; it will be enough for the present purpose to observe, that the acquired 

 tendency to read such undistinguished elements into meaning must be very 

 deeply fixed ; to all purposes, it might be considered as instinctive. For, while 

 all can at once see and designate an ordinary expression, which is the result of cer- 

 tain lines of feature, the artist alone can discriminate the characteristic curve, and 

 reproduce the effect on his canvass. But now observe the consequence of the 

 associating tendency, — the strong prepossession which conveys ideas of expression 

 from lines Indistinctly discerned, will actually select and attach similar expression 

 to similar lines, when they appear in any mass of confused and indiscriminate 

 lines. The instant the eye rests on a single characteristic curve, this will be the 

 key to all the lines in the mass which (if I may so speak) belong to the same face. 

 The fierce eyebrow will impose on the eye a mouth of the same character, which 

 will be seen in its proper place. This case is the plainest of its class ; but all 

 the forms of familiar things are similarly traced by the vacant eye, out of formless 

 elements ; for these alone leave it free to the stream of association. From this, 

 I might proceed to the phenomena of dreams ; but the subject demands a separate 

 treatment, and must be referred to the conclusion of this Essay. 



The cases so far stated to the Academy have exhibited the simple continua- 



* The picture alluded to is the " Arran Fisherman's Drowned Child," by Burton. 



