The accelerated implosion effort, which began in August 1944, made rapid progress. By October 1944, James Conant was giving a lensed implosion device a 50–50 chance of working on schedule, if all went smoothly, for a test at Trinity on 1 May 1945 and a “3:1” chance for a test on 1 July. But he added, “In my opinion, the probabilities of success by the gun method (Mark 1) within the next year are very much greater than by the implosion method. Indeed the gun method seems as nearly certain as any untried new procedure can be.”

Overcoming asymmetries remained the outstanding technical problem of the implosion program. By mid-fall 1944, two experimental strands of the implosion program were converging on this problem: research on the explosive lens and on the electric detonator. In addition, in T-Division, Robert Christy put forth a conservative proposal for overcoming the asymmetry: try to implode a solid sphere rather than a spherical shell of active material. However, calculations indicated that the “Christy gadget” was intrinsically far less efficient than the hollow weapon, and that such a device would require a modulated initiator to activate the explosion at the most favorable moment. The call for the development of the implosion initiator added another thorny problem to the program.

As the time approached when sizable quantities of plutonium would become available, gross design features had to be frozen in order to begin final bomb production.