220 
THE JOURNAL OF BOTANY 
In. rupestribus cretaceis ad Malko Tirnovo in monte Strandja, 
Bulgaria austro-orientalis. 
This new Veronica is easily distinguishable from the other species 
of the group to which it belongs, as follows: 
V. Teucrium L.—Perennial but herbaceous. Stems erect, tall. 
Stems hairy all round. Leaves soft, membranous and pilose, oval or 
elliptic-cordate, sessile 12-22 mm. broad. Sepals glabrous or ciliate 
at the margin. Fruit pilose. 
V. prostrata L.—Perennial but herbaceous. Stems ascending or 
prostrate. Stems hairy all round. Leaves rugose, membranous and 
pubescent, elliptic-oblong, shortly petiolate, 4—(5 mm. broad. Sepals 
glabrous. Fruit glabrous. 
V. rhodopea (Vel.) Degen.—Perennial but herbaceous. Stems 
ascending or prostrate. Stems hairy only along two opposite lines. 
Leaves rugose, membranous and glabrous, only ciliate at the margins, 
elliptic-oblong, shortly petiolate, 4-6 mm. broad. Sepals ciliate at 
the margin. Fruit slightly puberulous at the summit only, otherwise 
glabrous. 
V. Turrilliana , sp. nov.—Suffrutescent. Stems ascending or 
prostrate. Stems hairy only along two opposite lines. Leaves coria¬ 
ceous, persistent, quite glabrous, oval or elliptic-oblong, shortly petio¬ 
late, 6-8 mm. broad. Sepals glabrous. Fruit slightly puberulous at 
the summit only, otherwise glabrous. 
It is noteworthy that V. rhodopea and V. Tarrilliana, both 
endemic to the Bulgarian flora, are quite isolated plants ; no inter¬ 
mediates are known which connect them with one another or with 
what are apparently their closest allies— V. prostrata L. and V. Teu¬ 
crium L. : at the same time intermediate forms between V. prostrata, 
V. Teucrium, V. austriaca , and other species of this group, do occur 
in Bulgaria. It is also worth recording that our plant and the species 
compared with it above show ecological differences in the habitats 
respectively favoured. V. Teucrium and V. prostrata are widely 
distributed over all Bulgarian territory, as in many other parts of 
Europe; the former is common in the dry oak and mixed forests, and 
the latter is not rare in the dry pastures of the plains. On the other 
hand, V. rhodopea and V. Turrilliana are apparently narrowly limited 
in their distribution ; the former is a West Rhodopean endemic, 
occurring in moist mountain pastures, while the latter is a xerophyte, 
restricted to the small portions of cretaceous rocks which remain 
oasis-like among the large forests of Vagus orientalis and other trees 
of the Strandja range. The isolated position of these rocks in the 
forests is connected with the complete absence of a specific xerophvtic 
vegetation. With the exception of V. Turrilliana and Sideritis 
taurica M. B., all the other forms occurring in these habitats are 
common Mediterranean therophytes, apparently of relatively recent 
introduction. It is thus possible that our new plant is a descendant 
from V. prostrata and was derived from this species, in an older 
epoch, through the agency of the habitat conditions, and is not, there¬ 
fore, a recently formed ecological race. 
