Document R2Lxg3YyV7D61m7gvXzg71JZE

October 5, 1540 Pr Graham Edgar Dear Dr. 'dgart ins following are my cements on your letter of October 7 regarding the iinay burner for bthyl Oasoline \ \ 1. Dquigmont * - It will be inheregtiTtg~~bcf so the detailed drawing, although the photograph is fairly, descriptive. xhe filter pads sumitted.seem rather densef and I aa wonder ing how much back press ure^tBoy put on the system, and 3\ow. h.ie back pressure Ijr' cvereoite* I would expect a thicker but looser pad to be Equally as effective, while showing less pressure drop.x 2. "cleulationg - I agree -with your remarks regarding tho amount of air and fuel Used. If the amount of lead actually presetti-^-hbrnbustlba gases is as small as re ported, tho'-boxiciusicns would not be unvalidated by a three-fold' correction factor, 3* The only question in toy raind relates to the accuracy of the values* If no load was found in the fifth absorber, there is still-the possibility that some load was present in h'Hcn-absorbabla form, I believe .that early experiments on the exhaust of an engine failed to account for all the lead input, end this was probably due to inability to remove tho lead quantitatively from the exhaust gases. It my be, however, that in this case.jse-'afro dealing with Ffa and not FbO, and therefore, that th&^rocovery is more quantitative. A roubh check on this would be obtained by analyzing a filter, which according to tho Army report^ should account for 99+% of the lead input (including, of* course, the sera p ings from the cracking tube). 4. Further tests. - An improvement in the testing technique Will" `consist' of analyzing the measured gas for CO end C0G, thus obtaining a measure in terms of fuel input. Ibis vrill remove tho necessity of gathering all the combustion K 00085! Dr* Graham Edgar Fage 2 Cot. 9, 1940 products, or of preventing admixture with fresh air* The advisability of running further tests, however, vri.ll depend mostly on rr* i&ohle's cpionion of the data sv&raitted, 5, A sample of ^bril't-C-lsme* hae been received and will be identified. Unless its ocmpeeit'ioxh should have something unusual, I doubt v/he'thQrtthl'S'-will be worth testing* . Very truly yours. ';c sAMC co: Dr. ^'.F.'laohla K 0009552