336 CREATIVE SCIENCE 



I have a threshold of impressionahility (hereafter, TI) that most 

 stimuh do not surmount. Two factors are involved: the "subjective" 

 setting of the threshold value and the "objective" nature and magni- 

 tude of the stimulus. Reflecting the first factor, some people can con- 

 centrate in the midst of distractions which deprive others of all pos- 

 sibility of fixing their attention. Reflecting the second factor, a loud 

 or abrupt noise breaks in on my attention when a softer or more sus- 

 tained noise does not. But what matters most is neither factor by it- 

 self, but rather the extent to which the stimuli do or do not fit what I 

 accept tacitly as the pattern of normality or non-significance. The 

 waxing and waning roar of my neighbor's mower makes no impres- 

 sion on me, though I would become aware of the faint scratching 

 at the door were I not the owner of a dog. 



The scientist in his laboratory has also a threshold of impression- 

 ability. In any experiment some things can be observed only at the 

 price of willingness not to "see" a multitude of others. The horizon of 

 expectation defines the focus of the scientist's attention. To events 

 beyond that horizon he gives little or no consideration, which is of 

 course why the peripheral class of accidental discoveries is so small. 

 But even within the focal area not everything will be noticed. The 

 scientist's training and experience teach him that in all experiments 

 there are apt to be minor variations arising from assignable and un- 

 assignable fluctuations in the experimental conditions. To be recog- 

 nized even as a discrepancy a focal datum must surpass some vague 

 margin of reasonable variation. This point is well displayed in the 

 very general scientific problem of identification. 



Consider how Priestley came to discover "dephlogisticated air"— 

 what we call oxygen. First comes the focal discovery that the gas 

 from mercurius calcinatus per se is a brilliant supporter of combus- 

 tion. The eflFect is large and striking, and immediately put Priestley in 

 mind of a "factitious air" he had earlier found to behave similarly. 

 The property was, as far as Priestley knew, unique. He had then no 

 hesitation in (mis) identifying the new gas as what we call "laughing 

 gas." Actually there are perceptible differences in the appearance of a 

 candle burning in oxygen and one burning in laughing gas: in oxygen 

 the combustion is somewhat more brilliant, and in laughing gas some 

 brownish vapors are produced. The differences are noticeable, but 

 they were not at first noticed by Priestley. These combustion phe- 

 nomena are not highly reproducible in experiments as crude as 



