102 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 54 



language. What I have designated as the natural method of 

 describing plants vindicates its right to that title by the fact of its 

 primitive universality. If I speak of it otherwise as the method 

 of comparisons, in allusion to what it chiefly exacts of him who 

 would use it: that is, first, familiar acquaintance with certain 

 specific types as standards of comparison, and second, the ability 

 to construct a mental image of the unknown by means of the 

 describer's telling in what several particulars it differs from the 

 given type known to both. By way of illustration, an untaught 

 woodman, familiar with junipers, reports to a botanist that in a 

 new region to which he has wandered there are trees in all respects 

 like red cedar, or juniper, except that instead of yielding berries 

 they bear round cone-like bodies approaching the size and form of 

 smallish walnuts. The botanist at once pictures in his own mind 

 cypress trees, and assures his informant that his new trees are 

 cypresses. Such is the method of comparison in phytography; 

 and it may quite as aptly be called the natural one, for it is that 

 invariably used by the primitive and untaught; not, however, 

 always very well and successfully. Being the primitive method, 

 it is therefore that of Theophrastus ; not, however, to the complete 

 exclusion of a number of absolutely definitive terms such as are 

 now in use and always have been. The Greek observed that, 

 taking the plant world as a whole, the leaf is the most endlessly 

 diversified of organs, and also that within the limits of a species 

 its form is constant; from which two conditions it follows that no 

 other organ is so readily available in making distinctions between 

 different plants. Now as to the art of plant-description by the 

 method of comparison, it is very necessary that the types to be 

 used in comparing be chosen considerately. There seem to be 

 indications of his having thought upon this matter, though he has 

 not explained, or even didactically set forth his scheme. It is only 

 by a certain order and fitness in the scheme itself that one infers the 

 author's having studied it out. The most common types of leaf 

 outline are perhaps the lanceolate, ovate, oblong, and suborbicular. 

 He knows nothing of any such terms; but when he wishes to say 

 that which would express what modern botany expresses by the 

 term lanceolate, he says the leaf is that of the laurel, i. e., Laurus 

 nobilis. A leaf that we should describe as oblong, unless it be 

 much too large, is with him that of the olive tree. A leaf that is 

 of rounded contour and nearly as broad as long is compared by 

 him to that of the pear tree. For the ovate in outline his type 

 is the leaf of the ivy, Hedera Helix, in respect to which one must 



