320 



T. G. TUTIN 



In using cytological criteria it is of course a great help if information is available 

 about the kind of relationship between the different basic numbers which sometimes 

 occur within a genus, as this may enable a decision to be made about their taxonomic 

 significance. 



Chromosome size is generally of more value as a guide to larger classificatory units 

 such as tribes, subfamilies or families, than as a generic criterion, though it may indicate 

 genera which require investigation anew. For instance Brachypodium is reported to 

 contain species with large chromosomes and others with small ones, and has three different 

 basic numbers, 7, 9 and 15; it is therefore highly probable that once again superficial 

 similarities mask fundamental differences. 



It seems desirable, though possibly unnecessary, to point out once again that taxono- 

 mists using cytological data should do so with great caution. Many of the earlier chromosome 

 counts are quite misleading : for instance the first counts made on Glyceria (Stahlin, 

 1920) included the numbers 28 and 56 for G. maxima, while the true number is without 

 doubt 2n = 60. Some of the apparent discrepancies in chromosome number are almost 

 certainly due to mistakes in identification of the plants on which the counts were made, 

 mistakes which cannot. now be rectified as cytologists rarely keep specimens. 



In some families, for example Orchidaceae and Gramineae, a number of naturally 

 occurring intergeneric hybrids are known and others have been made artificially. This 

 situation seems to be regarded by most taxonomists as just one of those things which 

 sometimes happen, but which do not invalidate the genera concerned. Geneticists on 

 the other hand quite naturally regard the possibility of gene- exchange as of far greater 

 importance than any number of morphological dissimilarities. Failure of two plants to 

 cross may be due to a single gene or may, at the other extreme, be due to a totally different 

 genetical make-up. The sterility barrier, from whatever cause it arises, obviously has an 

 evolutionary significance, and it is with this that the geneticist is concerned. Unless, 

 however, sterility is accompanied by some perceptible degree of differentiation the taxono- 

 mist, who is concerned with classifying organisms as they exist today, cannot be expected 

 to take note of it. 



The question of the treatment of genera between which some hybridisation is possible 

 is bound up with the fundamental assumptions on which our scheme of classification is 

 based. This scheme is basically a morphological one, on to which varying amounts of 

 phylogenetic speculation (it cannot be called fact) have been grafted by different workers. 

 At the species level solid facts about phylogeny are sometimes available. For genera and 

 higher taxa definite unambiguous information about evolutionary relationships is rarely 

 obtainable and there seems no great prospect of getting it in the future. 



It seems impossible to make a firmly based phylogenetic classification, and even if 

 we could do so it would probably prove highly inconvenient, though interesting. In view 

 of this I think that, while we should pay due attention to what genetical evidence is avail- 

 able, we should not utterly disrupt parts of our classification to incorporate it, while the 

 major portion has to remain on a morphological basis. 



Turning once more to the Gramineae for examples, in the tribe Hordeeae Stebbins 

 and his co-workers have made a series of intergeneric hybrids connecting up all or nearly 

 all the genera so that, from the genetical point of view, the numerous and diverse species 

 included in this tribe should be regarded as forming a single genus. This seems to me to 

 be an unjustifiable conclusion. What Stebbins has shown is that the plants included by 

 taxonomists in the Hordeeae form a natural group related phylogenetically to one another. 

 During the course of evolution changes have occurred which have produced a considerable 

 degree of morphological and ecological diversity but which do not happen to have given 

 rise to sterility barriers. The absence of these barriers is of significance from the point of 



