NATURAL HISTORY, STATISTICS, AND APPLIED MATHEMATICS 249 



went and lay under the big oak until his fellow students solved the problem 

 by their own slower-witted techniques. He is today a highly successful pro- 

 fessional rose breeder, but as I follow the careers of my other students I note 

 that they all profit by the methodology of Natural History, though there is 

 only one of them who really works in that field. I note in particular that it 

 is used in their everyday thinking by two or three who, when they were 

 graduate students, had no great gifts in that direction or had, by their under- 

 graduate mentors, been taught to despise such methods. Yet there is no better 

 way for getting to grips with the fundamentals of a big problem than the 

 method of Natural History. The problem is at first a set of unrelated facts; 

 as the trained naturalist examines it, he perceives groups of facts organized 

 into patterns, patterns recognized because they repeat their essential features 

 time after time. 



One of the outstanding points about Natural History is the large extent 

 to which it deals with pattern data, rather than with pointer readings, 

 lengths, widths, densities, weights, etc. To be specific, in the problem I set my 

 class, the barbed-wire scars were a significant pattern, the trees coming up 

 in groups of two or three as stump sprouts were another, the dead and dying 

 branches which once extended out from the uncut forest to the cutover 

 pasture were still another. An experienced observer could use one set of pat- 

 terns to set up a hypothesis, then figure out the logical consequences of this 

 hypothesis and check with other data to confirm it. 



One reasoned something as follows: fence scars in a line, and one of them 

 has a piece of barbed wire; this must be a boundary between two fields. Ah 

 yes, the stump sprouts are all on the east side, and there are none on the west. 

 If the east side was cut over, then the trees along the fence should have sent 

 out branches sideways into the light and there should be either the scars of 

 such branches or the branches themselves more or less dying off in the reduced 

 light. A quick glance confirmed this hypothesis; a checkup on the one-sided 

 distribution of light-loving plants, pasture roses, red cedars gave added sup- 

 port. One was almost as certain as if one had been present that this was 

 indeed an old fence line between two fields, one of which had been cut over 

 and the other not. One's mind and eye worked back and forth from data to 

 hypotheses, then back again to data. Inductive and deductive reasoning 

 tripped so closely on one another's heels that they were one simultaneous 

 process. 



Confronted with any large and complex problem, in any field, the scientist 

 who has had effective training in Natural History knows more or less in- 

 stinctively what to do. Everything looks chaotic at first, but we do not live 

 in a chaotic universe. There may be confusion in our minds, but there is no 

 chaos in the way the world is running. Faced with such a problem, the 

 properly trained scholar looks around for significant repeatable patterns in 

 the data and reasons back and forth from observation to hypothesis until he 



