SCIENCE. 



[N. S. Vol. VIII. No. 208. 



our inability to deduce anything from a 

 single experience. The burnt child may 

 dread the fire as much as if it had been 

 burned twenty times, but the only way for 

 it to learn whether, and to what degree, its 

 dread is wise and prudent, without passing 

 through the slow and painful process of se- 

 lection, is to get knowledge, for a single ex- 

 perience affords no basis for any logical 

 process. 



While the emotional value of a sensation 

 is, no doubt, limited by inherited structure, 

 and dependent, to some degree, on inten- 

 sity, its objective value as knowledge is 

 regulated in accordance with the statistical 

 law of probability. 



If the history of what we call our uni- 

 verse were complete from beginning to 

 end ; if everything which exists in it were 

 reduced to mechanical principles, and 

 traced back to primitive nebulosity, this 

 history would be only a single experience 

 in cosmogeny, so far as the history of uni- 

 verses is in question. If we were to find, 

 somewhere, a second nebulosity we would 

 not be able to infer anything, except from 

 the worthless analogy of a single experi- 

 ence ; nor would we be able to infer or de- 

 dace, from our own, anything, not already 

 known, with more than reasonable con- 

 fidence. If we were still ignorant of any 

 part of our order of nature we should have 

 no way to find out but the way we have 

 now, and while our confidence in its stabil- 

 ity would be reasonable and judicious it 

 would not be necessary or absolute unless 

 our experimental knowledge were also ab- 

 • solute. 



It seems to me that the truth for which 

 Huxley strives, and hits with imperfect 

 aim, would be more correctly expressed by 

 the statement that, if our knowledge of na- 

 ture were to be made complete, from begin- 

 ning to end, we should expect to find that 

 our confidence in its stability had been rea- 

 sonable and judicious, and wise through- 



out, and that any other expectation would 

 have been folly and suicide, bodily as. well 

 as mental, and that it is only in this sense 

 that we could assert that it all lay poten- 

 tially in the cosmic vapor. 



It is not because I dread or fear the phil- 

 osophy of evolution that I refuse to accept 

 it, but because it is not yet proved. When 

 it is proved I shall accept it with cheerful- 

 ness, for I most assuredly hold no belief 

 which is inconsistent with it, although I 

 fail to see how the reduction of all nature 

 to mechanical principles could show that 

 nature is determinate ; for if exhaustive 

 knowledge of ' primitive nebulosity ' should 

 sometime show that there is nothing in na- 

 ture which might not have been expected, 

 I cannot see how this could show why the 

 things we expect should be the things which 

 come about. 



They who assert that complete knowledge 

 would be fore-knowledge forget that, for 

 minds like ours, the only source of knowl- 

 edge, either complete or incomplete, is evi- 

 dence ; for evidence can tell us only what 

 has happened, and it can never assure us 

 that the future must he like the past. Even 

 if we knew all that has happened, from the 

 beginning down to the present moment, we 

 should have to regard the unknown re- 

 mainder as equal, in all probability, to the 

 known jjast. To my mind, Jevons' demon- 

 stration that if certainty be represented by 

 unity the utmost confidence we can ever 

 reach by complete knowledge can never ex- 

 ceed a value of one-half seems conclusive ; 

 but even if it be increased until it differ 

 from certainty by less than any assignable 

 quantity it must still remain nothing but 

 reasonable confidence. 



There may be some unknown reason why 

 the stone which I set free from my hand shall 

 fall, and it may be that, as my mind has been 

 shaped by natural selection, I am unable 

 to expect anything else than that it shall 

 fall ; but science affords no evidence that its 



