Dear Josh, i ! August, 29, lgtjg , I'm sorry for the delay in answering your letter. I've been on vacation. *hanks very much for your interest in and your comments on the arabinose paper. b'irst, to answer some of your queries. 1 have oompleted the enzymatic analysis of all the mapped arabinose mutants. However, it takes me so long to write a presentable paper that 1 thought it best that the genetic and physiological experiments with intact cells be published now, while it may be of some interest. byway, I have so often been asked by editors to cut long papers that 1 decided to save them and myself the trouble of re-writing. L-arabinose induces both isomerase and kinase. L-ribulose cannot be used by intact oells of B/r wild type or mutants. Group c may be permaase negative. &wever, 1 have been unable to force induction of either the isomerase or kinase by growing the cells in high concen- trations of arabinose. We have not tested for dpimerase activity. Group C may be deficient in the epimerase, as well as in isomerase and kinase. As 1 mentioned in the footnote(5), Group B mutants, although all kinase negative, have different levels of isomerase activities, varying from l/10 to 4 times the isomerase activity of the prototroph. 1 am aware of Kalckar's and %ther's work. However, since the purpose of this paper was to describe a reliable methodology for ordering closely linked markers and to indicate the general unreliability of fiemerec's and fiartman's method, 1 did not think it was necessary to mention the galactose work. As far as the tables are concerned, you must remember that we are only interested in drawing attention to that data bearing on the order of the ara sites in the simplest possible manner, on the basis of the reciprocal crosses. 1 think anyone reading the paragraph entitled "brder of ara sites" on page 8 and referring to Fig. 1 and working out a few crosses with a pencil and paper should be able to grasp the method.