1042 REPORT— 1898. 



plants as a whole, and tliat of the classification of their parts. In each caso the 

 earlier srstems were artificial. We may compare the Liunaean system of taxonomy 

 with the Ilofmeisterian organography : in both the rigid application of a precon- 

 ceived method placed incongruous things in juxtaposition, in each case a widening 

 of the basis of the classification has resultea in a redistribution on more natural 

 lines. The present ideal of taxonomy is the same as that of the phylogenetic 

 organography, viz. to group according to descent. The limitations are alike : 

 systematists and morphologists both find their greatest difficulty in the incom- 

 pleteness of the record, and the frequent isolation of the things to be classified. 



But without following the obvious parallel further, we may now briefly review 

 OUT position as i-egards organography, and the following categories are to be 

 recognised, though they graduate almost imperceptibly into one another : — 



Ilomoytrnj. — («) Repetition of the individual part in successive generations, 

 with the same number and position. This is exemplified by the cotyledons, the 

 foot, and first root. 



ifi) Essential correspondence of parts varying in number and position, but 

 corresponding in character and development, produced in a regular sequence ; e.g. 

 most cases of continued embryology. 



(c) Transferred position of parts, similar in origin and structure to those pro- 

 duced in regular sequence ; e.g. roots, adventitious buds, sori oiAspidium anomalum, 

 aposporous and apogamous growths, many monstrosities ; these we may believe 

 to result from a transfer of inherited developmental capability. 



Homoplasy. — This may be recognised with varying degrees of probability ; 

 starting from cases where the question of community of descent is open (as with 

 nearer circles of affinity), and proceeding to those in which distinct evolution is 

 virtually certain. It remains for future investigation to clear up doubtful points. 

 Meanwhile, taking the case of leaves for the purpose of illustration, we may con- 

 template the following possibilities : 



{a) A possible origin of two homoplastic series of leaves in the same plant, 

 and the same generation {Phi/Iloglossum). 



ib) Two homoplastic series in the same plant, but in different generations 

 {Lycopodium cernuum). 



(c) A possible distinct origin of homoplastic leaves in distinct phyla, but in 

 the same generation (sporophyte of Ferns, Lycopods, Equiseta). 



{d) A distinct origin of homoplastic leaves In distinct phyla, and distinct 

 generations (e.g. leaves of Bryophyta and of Pteridophyta). 



Now Homology has been used in an extended sense as including many, or even 

 all, of these categories. It seems plain to me that this collective use of the term 

 homology carries no distinct evolutionary idea with it ; it indicates little more 

 than a vague similarity ; the word will have to be either more strictly defined or 

 dropped. The old categories of parts based upon the place and mode of their 

 origin are apt to be split up if the system be checked by views as to descent. 

 Comparison, aided by experiment, supersedes all other methods, and the results 

 which follow raise the question of terminology of parts which have arisen by 

 parallel development. 



In parts which are of secondary importance, such as stipules, pinnre, the 

 indusium, hairs, glands, the inconstancy of their occurrence points to independent 

 origin by parallel development in a high degree ; in parts of greater importance, 

 such as leaves, a parallel development may also be recognised, though in a less high 

 degree ; in the case of sporangia their acceptance as a category sui generis dis- 

 pelled the old view of their various origin from vegetative parts ; but we must 

 remember that this does not by any means exclude a parallel development also in 

 them, by enlargement and septation from some simpler spore-producing body, 

 though this is not yet a matter of demonstration. Finally, the sexual organs 

 are probably homogenetic in all Archegoniate plants, but we have no proof that 

 sexuality arose once for all in the lower plants ; the probability is rather the con- 

 trary. Thus we may contemplate as very general a polyphyletic origin of similar 

 parts by evolution along distinct lines, but resulting, it may be, in forms essentially 

 similar. 



