Suppose, ior example, tnat the plant in question, of which 
it is desired to learn the specific name, is a species of Rubus. Its 
stems are erect and covered with straight prickles, its leaves are 
compound, with 3-5 leaflets, its petals are white and the calyx is 
notably glandular. By reference to the key to Rubus (p. ___) it will 
be seen that the first alternative presented there refers to the leaf 
habit. Inasmuch as the leaves of the plant being determined are conm- 
pound, we necessarily choose the sepnd alternative which requires Ge) 
this condition. Passing to the next pair of alternatives, to the 
next fork in this tortuous path, it is again necessary to choose the 
second alternative which requires a plant which is erect and armed 
with prickles. inasmuch as the petals are white, we choose again the 
second alternative of the next pair and pass to the last pair. pet 
plant must therefore be either R. ideeus or R. leucodermis. We must 
now choose between the last alternatives, and are constrained by the 
straight prickles and glandular hairs of the calyx to halt at 5. R. 
ideeus. Turning to the descrivtion or that spvecies, we compare our 
plant with it and find-that it compares favorably, that is, fits within 
the range of variation described. Thus we conclude that the plant in 
question is that svecies. 
Sometimes the key may be inadequate or 
ex, or it may 
be applicable only to the vlants in Slower, and the specimen we have at 
w 
hand ig in fruit. It is still possible to arrive at a conclusion by 
orocesé Oo: elimination. For example, let us suppose that the plant in 
question is an Eriogonum in which only some weathered “ruits ere AVALLA 
