GENERAL CONSIDERATIONS 23 



extent, a matter of training and experience. 

 One man's descriptions are apt to be consid- 

 erably better than another man's, and the 

 one who consults the written description has 

 a right to know its value as nearly as pos- 

 sible. Certain points in description, more 

 over, as elsewhere explained, are always 

 matters of purely personal judgment, and 

 the person who renders the judgment should 

 therefore be held responsible in the record. 



The date at which the description is made, 

 is also a matter of some interest (sometimes 

 of real consequence), though its value is more 

 bibliographical than descriptive. It should 

 always be given. 



Before we pass now to the description of 

 the fruits themselves, we may stop for a 

 single observation regarding the nature of 

 the sample to be chosen. It goes without 

 saying that the specimens chosen for descrip- 

 tion should be representative typical. It is 

 desirable in many cases, of course to preserve 

 notes of abnormal samples, but the fact of 

 their abnormality should be conspicuously 

 advertised in the description. In all ordinary 

 cases great pains will be taken to select only 



