254 THE JOURNAL OF BOTANY 



None of the type had invaded this particular area, which was 

 thickly covered with the flowering-stems, mostly as far prostrate 

 as the grassy sward in which it grew allowed, while some were 

 ascending, and from 10-13 cm. high. The usual down-land form 

 was about 10 cm. high. The two plants do not form so striking a 

 contrast in the herbarium owing to the loss of colour in drying. 



The leaves of both forms are ovate-reniform and deeply crenate, 

 but there is some difference in the crenations ; in the type they 

 are rounded and more cut in at the base, the central one having a 

 rounded apex sometimes terminated by a mucro ; in the variety 

 they are more pointed and the central one is acuminate, without 

 a mucro. 



With regard to the hairiness of the plant, there is a great 

 difference between the two. The type has a short pubescence on 

 the stem, composed of simple hairs with three or four cells ; in 

 the variety the hairs are longer, containing about eight cells, 

 while the nodal tufts have about sixteen cells, and in the type 

 about ten. The typical down-land plant has dark green or 

 purplish green leaves, with a polished surface and short scattered 

 hairs of one to two cells ; in the variety the leaves are grey-green, 

 with a dull surface, and are covered with long bristly hairs with 

 six to eight cells, which give the leaves the soft hoary appearance. 



With reference to the flowers, the calyx in the type is 5 mm. 

 long, in the variety 4 mm. The corolla of the type is 16 mm. 

 long, with a longly exserted tube, and the colour is purplish blue. 

 The corolla of the variety is only 8 mm. long, and the corolla- 

 tube is included in the calyx ; the number of flowers in a whorl 

 is more numerous usually than in the type, and the colour is pale 

 lavender. The variety is apparently a female form, for stamens 

 were absent from all the flowers examined. It is curious that the 

 difference in sex should be accompanied by so many variations 

 from the type in the vegetative characters, and should give the 

 plant a more important varietal standing than the remark in 

 Hooker's Student's Flora, "Flowers dimorphic, larger 2-sexual, 

 smaller female." In Babington's Manual, ed. 9, it is thus alluded 

 to, "A small subglabrous form with short corolla is var. parviflora 

 Benth. " ; while, as a matter of fact, the variety as I found it is 

 far more hairy than the type. 



The flowers were in too young a state to have ripe nutlets ; 

 but both in the type and the variety the young nutlets looked 

 exactly alike as far as comparative development went. No doubt 

 the flowers are pollinated by insects from the typical plants close 

 at hand. The plants were vigorous, and gave the appearance of 

 spreading from a centre. 



The variety is first mentioned in the London Catalogue in the 

 eighth edition, 1886, and is also placed in the ninth edition, 1895; 

 but in the tenth edition, 1908, it is no longer found. What may 

 be the reason of this ? It is a very marked plant, surely more so 

 than many described varieties, and was evidently thought to be 

 so by Bentham, who described and named it, and who was usually 

 averse to the splitting up of species. 



