100 



Mr P. Sewell on the Morphology and [skss. li. 



Perhaps the most noticeable divergence from the normal 

 bilabiate condition common throughout the order is the four- 

 parted condition in Veronica. 



This has been held by most of the older botanists as an 

 ordinary case of tetramerous construction. 



Indeed, most recent systematists mention the occurrence of 

 4 or 5 lobes in the corolla for the purpose of cataloguing the 

 genera. Payer, Henfrey, and others have shown how this con- 

 dition is brought about, namely, by the union of the two upper 

 petals into one, and by the suppression of a sepal ; by this is 

 shown the true relationship of Veronica to the rest of the order. 



Corolla of Veronica Buxhaiimii, showing two bundles in upper petal. 

 (The stamens are bent forward.) 



Payer* made, in connection with the varialjly pentamerous 

 or tetramerous condition in Limosella, the generalisation that 

 apparently tetramerous flowers have evidently resulted from 

 pentamerous types where they occur with sepals arranged 

 diarjonally to tlie axis, iKjt orthogonally, as in Cruciferte and 

 Onagracea-. 



After finding this conclusion of Payer's, I made some 

 examination of a limited imndjcr of species of Veronica, that 



* " Scrophulariaccte," 7''t«/u7fc6' Xuturclles. 



