THE EVOLUTION OF SCIENTIFIC THEORIES 281 



Stance, given the complete unobservability of the stellar parallax that 

 is the one "certain" consequence of the moving earth he had postu- 

 lated, Copernicus and his followers simply added to his theory the 

 drastic assumption that the stars are so immensely distant that their 

 parallactic displacement is immeasurably small. And we say this was 

 indeed the assumption that Copernicus (and everyone else) should 

 have made. 



Is the situation clearer when strain develops from the discovery of 

 entirely new kinds of data? The X-rays "accidentally" discovered by 

 Roentgen seemed for a time to strain classical electromagnetic the- 

 ory, but ultimately found snug accommodation within the theory. 

 The photoelectric eflFect accidentally or incidentally discovered by 

 Hertz ( in the celebrated experiment by which he seemed to put that 

 same theory beyond all possibility of doubt) we see in retrospect to 

 have been a far more ominous anomaly. Yet at the time nobody saw 

 it on the horizon as that portentous cloud no larger than a man's fist. 



Even strain that develops as or from apparent antinomies within 

 the structure of the theory may for long go unrecognized as such. 

 The luminiferous ether, it transpires, must be more tenuous than the 

 most rarefied gas, yet also more rigid and elastic than the strongest 

 steel. Today we feel that 19th-century physicists erred in not rec- 

 ognizing diis anomalous combination of properties as a symptom of 

 strain. We, on the other hand, regard as only a passing difiiculty the 

 various infinities that crop up so awkwardly in quantum field theory. 

 Will history approve our judgment? 



Where there are but a few anomalies to be pronounced as such, or 

 dissipated with supplementary assumptions, nobody may detect the 

 presence of strain. A few sensitive souls may react to any greater 

 accumulation of anomalies, if they are recognized as such. That re- 

 mains ever a large if: even in our own day few indeed felt anything 

 fundamentally amiss when a dense overpopulation of transuranic 

 elements had to be hypothesized to account for the results of neu- 

 tron-bombardment of uranium. Yet always by some the sense of 

 strain is felt sooner or later, and sooner rather than later when dis- 

 crepancies are striking and/or numerous. At that moment occurs the 

 decisive transition Bartlett sketches with an apt example. 



Some infectious diseases are bacteria borne; more are bacteria borne; 

 very likely all are bacteria borne. But no; here is one, the mosaic dis- 



