270 BIOCHEMICAL SYSTEMATICS 



substances, was valid within limits. However, he assumed the rather 

 tenuous position that a higher molecular weight indicated a more 

 complex substance. This idea has been attacked by Gibbs (1958) 

 particularly with respect to the alkaloids, which may be in some cases 

 low order polymers (for example, bisbenzylisoquinolines). 



Specifically, McNair (1934) attempted to correlate the serial 

 numbers of families of the Engler and Prantl system with the molec- 

 ular weights of their alkaloids, specific gravity of their essential oils, 

 and degree of unsaturation of their fats to support his thesis that 

 more advanced families produce more complex substances. Despite 

 relatively meager data, only slight positive correlation, a tenuous 

 basic assumption with respect to what constitutes true chemical com- 

 plexity, and a circular argument to begin with, he nevertheless later 

 concluded (1935), on the basis of these criteria that: 



(1) Herbs evolved from trees. 



(2) Monocots are more primitive than dicots. 



(3) The woody Magnoliaceae gave rise to the herbaceous 

 Ranunculaceae. 



(4) Polypetaly is more primitive than gamopetaly. 



(5) Many carpels preceded few carpels. 



(6) Apocarpy preceded syncarpy. 



(7) Some aspects of the Bessey system are superior to the 

 Engler and Prantl system, and some are not. 



Although a number of the points listed above may actually 

 be correct, the new evidence brought to bear on the questions by 

 McNair will, in the final analysis, be judged as of the most trivial 

 sort— if indeed it has any relevance whatsoever. It is possible that this 

 rather uncritical application of biochemical information had an ad- 

 verse effect upon the field, despite McNair's zealous interest in its 

 development. Some thoughtful systematists may have concluded from 

 these contributions that biochemistry had little to offer. 



More recently, Gibbs (1945, 1954, 1958) has been particularly 

 associated with efforts to enhance the general appreciation of bio- 

 chemical systematics, along with Hegnauer, whose work has previously 

 been discussed in other sections. Gibbs has not exaggerated the im- 

 portance of the biochemical approach but rather has discussed this 

 approach as only one of several to questions of phylogenetic relation- 

 ships. In his own investigations, Gibbs has limited himself to a few 

 relatively simple chemical characters, and it appears that in some 

 cases these are not among the most fruitful. Some characters he has 

 used are the presence of catechol tannins, presence of cyanogenetic 



