TriE t>LAN BOOItS 



the man who made the records. In some 

 cases the data of the Santa Rosa books are 

 even more minute and particular than those 

 of the larger tests. 



Mr. Burbank has a good many such books 

 as these, covering the experiments of many 

 years, embracing many thousands of words of 

 notation. For some years when he was 

 struggling to make both ends meet, he tested 

 seeds for eastern dealers, receiving ten cents 

 for each variety tested. This was work re- 

 quiring accuracy and record of the strictest 

 type: like his records of after years, it was 

 scientifically and commercially exact. 



It will be seen, the more closely one studies 

 the scope and sweep of this great work, that 

 accuracy of record on essentials is imperative. 

 A single error in this would throw out of gear, 

 so to speak, the whole machinery of a test. 

 The creator of the new fruit or vegetable or 

 flower would be utterly unable to tell whether 

 he was proceeding upon definite lines or 

 running through a whole series haphazard, 

 intermixing everywhere into other tests and 

 rendering the whole invalid. First and above 

 all, in a work of breeding carried on upon a 



381 



