ATOMIC ENERGY — JOHNS 185 



captured by U^^ without fission, (3) be absorbed by impurities, (4) 

 cause fission of another U^® nucleus. To get a chain reaction the 

 number of neutrons in (4) decreased by the sum of those in (1), (2), 

 and (3) must at least equal the original number put in. The first 

 three possibilities must then be reduced to a minimum. The first loss 

 was easily handled by making the pile of uranium being treated large 

 enough. When the edge of a cube is tripled in length, the surface is 

 increased ninefold, but the volume is increased twenty-sevenfold. So 

 the chance of escape through the surface per unit volume is much less 

 than it was before. From theoretical considerations Fermi calculated 

 how large, under attainable conditions, the pile should be and it 

 worked — a marvelous achievement. Losses (2) and (3) were much 

 more difficult to reduce to a minimum. Only the tremendous resources 

 of the United States could make possible their solution, and a billion- 

 dollar plant was built at Oak Ridge, Tenn. Since U^^^ and U^^^ are 

 chemically indistinguishable they must be separated by physical means 

 from the fact that U^^ is the slightly heavier of the two. Thermal 

 diffusion would separate the lighter fluid, U^^, just as our mothers got 

 the lighter Devonshire cream on a pan of milk. Gaseous diffusion 

 through barriers allowed the lighter IJ-^^ gas to pass more readily'. 

 The cream-separator trick of the centrifuge brought the lighter IJ-^^ 

 to the center and a huge electromagnet could deflect lighter ions more 

 than the heavier ones. Since some substances, for example cadmium, 

 soak up neutrons like a sponge, the problem of obtaining pure ma- 

 terials was also a formidable one. However, these difficulties were 

 sufficiently overcome and a pile was built in the squash court in Stagg 

 Field at the University of Chicago. Cadmium rods were inserted as 

 safety devices and on December 2, 1942, the first self-maintaining 

 nuclear chain reaction was initiated by man, even with the cadmium 

 rods only partly withdrawn. For the well-being of Chicago the pile 

 was torn down and moved to the Argonne Woods, some 40 miles 

 outside Chicago. It should be noted that there are always enough 

 stray neutrons about to start the chain reaction, so that no "match" 

 is required. 



MODERATORS 



It was early suggested by Fermi, Compton, and others that slow 

 neutrons, that is thermal or low-energy neutrons, could be used to 

 split U-^5 and that such might not be absorbed by U^^. Since the 

 neutrons coming out of the fission are always moving at high speed, a 

 search was begun for so-called moderators which would slow up such 

 neutrons before they were absorbed by U^^. If such could be found, 

 the separation of U"' from U^* might not be necessary. Now fast 

 neutrons are hard to stop, for they can pierce several inches of steel 



