Recent Progress Scientific Computing Journal by Comrie Leslie John (1 results)

Soft cover. Condition: Very Good. 1st Edition. Two single journal issues in printed wraps; 274 x 198mm; pp 129-148, article at 129-135 (August), and 205-220, letter at pp. 218-219. SCIENTIFIC COMPUTING PERFECTED One of the most substantial treatments of the use of everyday office machinery in 'scientific computing', only matched in detail in Comrie's own writings, in particular his 1946 article for Mathematical Tables and Other Aids to Computation. Comrie begins with the Differential Analyzer - perhaps the most obvious competitor for his own practice of adapting office machinery to scientific ends. A very revealing section of the paper deals with 'The Training of Computers', i.e. the ways in which human 'computers' will have to adapt to the new machine methods: "[Computing] is often given to or falls to the lot of people who by their very temperament do not take kindly to it if it gets beyond the slide rule stage, or takes more than five minutes. Among these people I include mathematicians, engineers, draughtsmen and, I am sorry to say, even some physicists." From here Comrie moves on to pen-and-paper integration, the method of finite differences, double-entry interpolation, the use of punched card machines - here referring to his famous calculation of the Moon's motion down to the year 2000 - as well as the solution of simultaneous equations and finally the construction of mathematical tables. One interested reader of the article was R.R.M. Mallock, of the Cambridge Mathematical Laboratory, whose machine for simultaneous equations was rather rudely dismissed by Comrie. The December 1944 issue contains Mallock's response to Comrie, followed by an apology by Comrie, in which he states clearly his preference for mass-produced over one-off machines - a theme that to recur throughout subsequent decades. Very good condition: stamped ('stack') to front cover but otherwise unmarked; stitched as issued; light cover marking and edge-wear; paper flaw to pp. 133-134 of the August issue, affecting 5 or 6 words. References: not in OOC, but see nos 253ff.