1887] — Lustruction in Geological Investigation, 819 
is asked to consider two simple alternatives; the large dike 
_ either was or was not moved on this fault-plane : what would be 
_ the consequence in either case? what test may be invented to 
_ settle the question? If the dike were not moved on this fault- 
_ plane (that is, if the dike were younger than the fault), its margins 
_ would be in line with each other on both sides of the fault; if 
_ the dike were moved on the fault-plane, its north and south 
_ Margins would be offset, and both by the same amount. The 
3 shaded areas marked X, Y, Fig. 8, are thus seen to be of critical 
_ importance. If X consist of dike-rock and Y of slate, then the 
_ first alternative is true; if X be of slate and Y of dike-rock, then 
the second alternative is true. Especial attention is called to 
_ this deductive method of planning out lines of work. It is not 
a sufficient simply to go out-doors and look about; the geologist 
_ Must look first, then think, infer, plan, and at last lead his obser- 
_ vations to the precise spot where they will do the most good; 
= Rot the most good to the theory that may from some accident 
3 have secured his preference, but the most good in deciding 
= critically among all the possibilities of the case. I think the 
q history of geology need not be read far before one finds un- 
fortunate examples of the neglect of this method. A fact is 
__ “oted and a suggestion offered : the suggestion fits, and therefore 
3 t5 accepted. But the student must be made to see that such a 
method will not lead to safe conclusions. It is quite possible 
: that several processes may bring about the (apparently) same 
7 "sult. In such case it is manifestly unsafe to conclude that 
= SY one process did bring about:the result merely because it 
3 oud, Independent evidence must be found to check and con- 
? mp Suggested process after its suggestion before any confi- 
4 dence should be placed in it; further, all other apparently possible 
5 ey must be ruled out before any one can be accepted as the 
| ios actual one. It is not easy to teach this to a class, and 
aes will slip through the work without much appreciation of it. 
‘ ins uence of events, as constituting a portion of geological 
| Sea is a matter of importance. The case just described 
Sa with an illustration of it. The large dike was faulted 
the intrusion of the smaller one, for on tracing the fault- 
‘the w, to the northern side of the quarry, it is found to intersect 
Sia member of the double dike, and brecciate it as well 
