64 SECTIONAL ADDRESSES. 



The problem is reduced to its simplest elements in the following 

 scheme: — 



/ Italics, 



f Lower-case Roman. 



F Capitals Roman. 



<f> Greek. 



Our genera are equivalent to the forms of letters :' Itahcs, Roman, 

 Greek, and so forth. The successive species are the letters themselves. 

 Are we to make each species a genus? Or would it not be better to 

 confess that here, as in the case of many larger groups, our basis of 

 classification is wrong? For the palaeontologist, at any rate, the lineage 

 o, A, a, a, is the all-important concept. Between these forms he finds 

 every gradation ; but between a and b he perceives no connection. 



; In the old classification the vertical divisions either were arbitraiy, 

 or were gaps due to ignorance. We are gradually substituting a 

 classification in which the vertical divisions are based on knowledge, 

 and the horizontal divisions, though in some degree arbitrary, often 

 coincide with relatively sudden or physiologically important changes of 

 form. 



This brings us to the last point of contrast. Our definitions can 

 no longer have the rigid character emphasised by Huxley. They are 

 no longer purely descriptive. When it devolved on me to draw up 

 a definition of the great group Echinoderma, a definition that should 

 include all tlTe fossils, I found that scarcely a character given in the 

 textbooks could certainly be predicated of every member of the group. 

 The answer to the question, ' What is an Echinoderm? ' fand you may 

 substitute Mollusc, or Vertebrate, or what name you please) has to 

 be of this nature: An Echinoderm is an animal descended from an 

 ancestor possessed of such-and-such characters differentiating it from 

 other animal forms, and it still retains the imprint of that ancestor, 

 though modified and obscured in various ways according to the class, 

 order, family, and genus to which it belongs. The definitions given 

 by Professor Charles Schuchert in his classification of the Brachiopoda 

 (1913, Eastman's ' Zittel ') represent an interesting attempt to put 

 these principles into practice. The Family Forambonitidae, for instance, 

 is thus defined: 'Derived (out of Syntrophiidae") , progressive, semi- 

 rostrate Pentamerids, with the deltidia and chiliJia vanishing more 

 and more in time. Spondylia and cruralia present, but the former 

 tends to thicken and unite with the ventral valve.' 



The old form of diagnosis was per genus et dijferentiam. The new 

 form is fer proavuni et modificationent. 



Even the conception of our fundamental unit, the species, is in- 

 secure owing to the discovery of gradual changes. But this is a 

 difficulty which the palaeontologist shares with the neontologist. 



Let us consider another way in which the time-concept has affected 

 biology. 



Effect of the Time-concept on Ideas of Relationship. 



Etienne Geoffroy-Saint Hilaire was the first to.compare the embryonic 

 stages of certain animals with the adult stages of animals considered 



