Vol. 65.] GLACIAL EROSION IjST NORTH WALES. 321 



hardly sufficient weight is ordinarily allowed by field-geologists, 

 namely, that in impartial scientific investigation, care and thorough- 

 ness are just as essential in the deduction of expectable consequences 

 from tbeory and in the confrontation of the consequences of theory 

 with the appropriate facts, as in the observation of the facts them- 

 selves in nature. Not only so : it is extremely important to deduce 

 consequences from theories and to confront them with actual facts 

 while one is still in the field, for otherwise the most critical and signi- 

 ficant facts may escape detection. One sometimes finds among one's 

 colleagues a disposition to set aside such deductions as have here 

 been presented regarding the pre-Glacial form of Snowdon and such 

 consequences as have here been worked out regarding the protective 

 or the destructive action of glaciers, as too fanciful to be seriously 

 entertained by those who are occupied in determining the actual 

 facts of earth-structure and form. With regard to this, however, it 

 may be fairly urged that the study of the earth to-day deals not only 

 with observations concerning the visible facts of present structure 

 and form, but very largely also with inferences concerning the 

 invisible facts of past history, and that there is no safe way of 

 determining the truth of the inferences concerning these past facts 

 that does not involve the deduction of the expectable consequences 

 of theory. 



The standard geological literature of the day teems with in- 

 ferences as well as with observations : not merely with inferences 

 of a subordinate value, but with inferences that are of fundamental 

 value in the science. The sedimentary origin of stratified rocks, 

 the disturbance of slanting or curved strata, the original continuity 

 of faulted strata, the organic origin of fossils, the lapse of un- 

 recorded time at unconformities, the metamorphism of crystalline 

 schists, and many other standard conclusions repeatedly met in 

 geological essays are not facts of observation, but inferences based 

 on facts of observation. Not only so : these and many similar 

 inferences are not of axiomatic verity ; they have had to win their 

 way to acceptance, and however well we may be assured of them 

 to-day, however closely their order of verity may approach that of 

 directly observable facts, they remain nevertheless only inferences. 

 It might be desirable to call these standardized results of geological 

 investigation by some such name as ' facts of inference ' in order 

 to indicate their origin and to distinguish them from ' facts of 

 observation.' But, however this may be, it is surely essential to 

 recognize that, whatever name the ' facts of inference ' may bear, 

 they have advanced from an original stage of speculation to the 

 assured position which they now occupy, only after submitting to 

 the most rigorous examination by at least a generation of geologists, 

 in the course of which the consequences of the original theoretical 

 suggestions have been repeatedly deduced and confronted with facts 

 of observation, while all sorts of less successful theories have been 

 tried and found wanting. 



The point to be emphasized is, therefore, that no theory or 



