1886-87.] Affinities of Scrophidariacece. 101 



I might obtain evidence in support of the conception other- 

 wise hinted at that it is virtually a pentamerous flower, 

 which has become specialised into its present condition. 



I found that in many species two bundles, or nerves, passed 

 into the upper petal, whilst each of the other petals had the 

 normal single bundle entering from the axis. 



Even an isolated instance of this would be very suggestive 

 as to the union of two petals into one; but, although not by 

 any means always present, the two bundles are so constantly 

 in the upper petal in different species as not to allow of any 

 other explanation. 



Of the species I have examined they are both present in — 

 V. Buxhaumii. V. gentianoides. V. pinguifolia. 

 V. rupestris. V. sjncata. V. Bendusoni. 



They, however, do not occur in the specimens of V. Lyalli, 

 V. maritima, V. longifolia, which I examined, but it is very 

 probable that complete union of the two bundles has taken 

 place, this havmg been observed as by no means an uncom- 

 mon occurrence in other cases. 



Monstrous flowers of Veronica often occur in which there 

 are five petals. This I have noticed in V. glauco-ccerulea, 

 Veronica spicata, and in a few other species. Also, Payer 

 mentions that it is normal for Veronica Teucrium to have five 

 sepals. 



Since making the above examination in regard to Veronica, 

 my attention has been called to a paper by A. Braun,* in 

 which he mentions Veronica, among many other plants, as an 

 example of an apparently tetramerous condition resulting 

 from a pentamerous. 



Comparison of the condition of calyx and corolla in Veronica, 

 thus established as falsely tetramerous, with the condition of 

 the outer floral whorls in less closely allied genera reveals 

 the interesting fact that Buddleia (placed by some botanists 

 in Loganiaceae), Lathra;a, and Plantago exhibit with their 

 apparent tetramery the diagonal arrangement of the sepals. 



Other points of resemblance among these genera are 

 emphasised by this fact, and Plantago may be referred to as 

 illustrating the fact that this relationship should be taken 

 account of in systematic works. Prantlt has asserted, on the 



* " Uber Pseudotetramere Bliithen," £o<. Zeit., 1875. 

 ^. . •; t Prautl and Vine's Textbook, p. 266. 



