54 BOTANICAL GAZETTE [janttary 



first, a study of the modes or phases of homoeosis helps us to estimate 

 at its true worth teratological evidence applied to the solution of 

 phylogenetic problems; secondly, the facts seem to throw light on 

 the method of evolution of some normal structures; thirdly and chiefly, 

 the facts of homoeosis constitute, as has been already pointed out in 

 the introductory passage, an important section of the data of morpho- 

 geneses. Let us examine these relations of the subject a little more 

 in detail. 



^ i. The study of homoeosis must somewhat increase the caution 

 with which we use deviations from the normal as aids to morphological 

 interpretation. 



abnormalities 



them the ground of phylogenetic inductions. The stereotyped 

 remark of writers describing monstrous specimens has been that such 

 aberrations "are very instructive "-an expression of faith either in 

 the phylogenetically reversionary nature of abnormalities, or in 

 the eternal inviolability of homology in morphogenetic sequences. 



Surmises 



the days when the evolutionary story of plants was less complete 

 than it is now, and when morphologists were driven to indirect and 

 speculative methods. The history of this subject is voluminous; 

 perhaps, though interesting as a study in speculation, it is more 

 valuable for the errnrc wi^vu :* r j- ., - . _,. jl 



embodies 



truth ; and it will not be entered into here 



Some 



formations 



., & ^ x ^uiiiicuions were used have been abandoned, 



and others have been settled by the discovery of direct evidence from 

 comparative morphological studies of living and „ti„rt r,1»nte. As 



maturin 



normal and abnormal forms, 

 morphological significance of 



many abnormalities falls to nil. For example, we now know that 

 he conversion of the nucellus of an ovule to a shoot, as in Aliaria 



ana in Mmcom ;~ i>_n ♦., - 7 



i . _ ~ A «■"* wuic lu a, snoot, as in nua* 



and ln Drosera, is wholly without phylogenetic meaning. From 



StUClV Of even nnrmol ^ _• ,• . ° 



a 



normal 



morphological categories; an 



if tv,^ u .. — ° J A "luipnuiogicai categories, ■ 



if the observations assembled in the present essay are accepted 



d 



in 



