TAXONOMIC VALUE OF MORPHOLOGICAL CHARACTERS 



13 



The graphic method is essentially a method of averages. Even 

 though it be true that the average graph of any species is relatively con- 

 stant this does not aid in the identification of non-average individuals. 

 We must deal with individuals and not with averages. 



The graphs are neither sufficiently distinctive nor sufficiently con- 

 stant to be of value in separating species. In those cases where two or 

 more species occur together on the same hosts the graphs will not, in 

 general, reveal the fact that more than one species are present. 



The range of observed variation in those species which have been 

 sufficiently studied is so great as to preclude the use of the graphs. 



In proof of these contentions I shall present certain specific instances. 



The two species, Pseudococcus crawii (Coq.) and P. quercicolus 

 n. sp. occur in the vicinity of Stanford University upon the same hosts 

 and have at times been found together. In life they are so very similar 

 that they are scarcely distinguishable, but structurally they are very 

 different, as a reference to the figures illustrating the two will show. 

 Text figures 1 and 2 present the graphs of these two species. These 



Fig. 1. Pseudococcus crawii (Coq.) ; antennal graph drawn from specimens taken 

 from Quercus chrysolepis in the vicinity of Stanford University. 



graphs were obtained from individuals taken from the same host plants in 

 the same vicinity. The two graphs coincide practically throughout and 

 it is obvious that no amount of contemplation of a graph, including both 

 species, would reveal the fact that two species are present. 



Again, let us consider the case of Pseudococcus ryani (Coq.) and 

 P. sequoiae (Coleman). These two species occur together on Sequoia 

 sempervirens in the vicinity of Stanford University and both have been 

 recorded from various species of Cupressus. The two species are suf- 



