RESEARCH ON XYLEM AND PHLOEM IO5 



of research comes only from a tremendous backlog of information and how 

 necessary it is in evolutionary studies to take broad samplings of the plants 

 involved. It will be sufficient in our case to tell something of the approach 

 to a study of this nature. This approach underlies much of the research in 

 morphology as a whole, even though, as Sporne of England has recently 

 emphasized, it is seldom acknowledged or described. 



Using statistical methods. Frost's approach to the problem of the origin 

 and evolutionary development of vessels emphasized two especially important 

 concepts. They are both based on assumptions, the validity of which must 

 be borne out by the harmony of the evidence when they are employed. One 

 concept, as expressed in words pertinent to our purposes, states that if 

 tracheids are more primitive than vessels of any type and are related to them 

 in a direct line, then the most primitive vessels are those most similar to 

 tracheids. The other concept, once again worded expressly for our purposes, 

 is that rates of evolutionary development of various features (such as length 

 and perforations) of vessels are correlated; that is, they occur at the same 

 rate. If in certain restricted types of plants — generally unusual in other 

 respects as well as in their secondary xylem — one or more features do not fit 

 into the major, statistically determined trends, these become exceptions and 

 illustrate minor tendencies, an example of which is the great length of vessels 

 in the stems of vines. Frost dignified these aberrations by using them as the 

 basis for a third concept called exceptions. It perhaps was necessary to so 

 emphasize these aberrations, because some investigators chance upon excep- 

 tional forms and, on the basis of these relatively minor irregularities, love to 

 dispute generalizations whose significance they never have really understood. 



Frost found that tracheids are characterized by great length and that by 

 association the longest vessel members are therefore most primitive. By use 

 of the methods of association and of correlation, he statistically determined 

 that among vessel members, the longest had narrow diameters, thin but 

 evenly thickened walls, long end walls with numerous transversely elongate 

 perforations, and scalariform pitting on the side walls. Hence all these fea- 

 tures are primitive in vessels. Frost furthermore showed that there was a 

 continuous gradation from these primitive, long vessel members to the short- 

 est members, which are characterized by heavily — but unevenly — thickened 

 walls, transversely placed end walls each with a single large circular per- 

 foration, and alternate pitting on the side walls. From these conclusions, as 

 well as other information, Frost concluded that primitive vessel members 

 with their long scalariform plates are derived from scalariformly pitted 

 tracheids and that the really fundamental difference between the two kinds 

 of elements was the loss of pit membranes in the end walls of primitive 

 vessel members. 



We have thus in secondary xylem a beautifully graded evolutionary se- 



