Early History of Gibberellin Research 469 



letter from me to Professor Sumiki in February, 1952, goes in part 

 like this, "In recent experiments we have failed to obtain even mod- 

 erate growth using your medium containing glycerol, dihydrogen 

 phosphate, and ammonium chloride. We have found, however, that 

 when magnesium sulfate is added, in even small amounts, growth is 

 greatly enhanced. We are wondering if you have found it necessary 

 or desirable to add magnesium sulfate to your fermentations." Final- 

 ly, a letter to me from Sumiki in March, 1952, "I am very surprised 

 that the sample of gibberellin A sent to you by my assistant while I 

 was in the United States showed little activity. The activity of that 

 sample was not tested but this is the first time we have heard of crys- 

 talline gibberellin losing its activity so fast. We are now preparing 

 a new sample to send to you." 



By May, 1952, these difficulties had been straightened out, and in 

 June we had a successful pilot plant run which yielded 12 g. of crys- 

 talline gibberellin. From this I was able to isolate by repeated crys- 

 tallization a sample of pure gibberellic acid with a rotation of -|-90° 

 (12). Later, we developed a chromatographic method that would effi- 

 ciently separate gibberellin Aj and gibberellic acid (11). Our experi- 

 ence was passed on to a number of fermentation companies in this 

 country, and before long there was enough gibberellin available here 

 for everyone. All this time, of course, the British workers were carry- 

 ing out their work on gibberellic acid. 



Dr. Brian: We started work either in 1951 or late 1950. Undoubt- 

 edly, the stimulus to us was the sudden spate of abstracts of Japa- 

 nese work which hadn't reached us during the war years. Looking 

 back on things, I am very much more struck by stupendous pieces of 

 luck that we had rather than by difficulties. I can just mention a few 

 of these. 



First of all, our strains of the fimgus, Gibberella fiijikiiroi, were 

 obtained originally simply by getting them from culture collections. 

 By far the best that we found in the early days and by far the best 

 that we still have was a strain which to my certain knowledge had 

 been kept in culture collections for over 30 years, and I believe, in 

 fact, was the type isolated by Sawada. We were very lucky indeed, 

 I think, to come across so stable an organism to work with. 



The second piece of luck we had was that in carrying out our 

 preliminary fermentations, we completely ignored any previous work 

 and used the kind of media that we had been used to using in our 

 other work on fungal metabolic products. We immediately got yields 

 of a gibberellin in far greater quantities than any previously-recorded 

 yields — again I think purely by accident (1, 2). Very shortly after we 

 got this material, one of my colleagues, Philip Curtis, said that this stuff 



