Notes. 
623 
purple tint and nearly always marked with striations (hence the name 
Z. striata , which is one of its synonyms) of a deeper tint, offer such 
obvious characters for differentiation. 
Up to the year 1890, only Z. vulgaris occurred about Oxford; it 
then showed little variability either in its flowers or foliage. In that 
year a space of ground between the Great Western and the London 
and North-Western Railways was filled up with chalk rubble brought 
from the chalk district of Upton in Berkshire. With the chalk a con- 
siderable number of plants characteristic of that formation was 
introduced. In that year a profuse growth of Iberis amara and 
Pier is hieracioides , and a less plentiful quantity of Linar ia repens } 
Campanula glomerata , and Bromus erectus, covered the chalk with 
a very different vegetation from that which occupied the other side 
of the railway. In that year no hybrids of Z. repens and L. vulgaris 
occurred ; but the following year a very large number of hybrid plants 
occurred, in fact they were sufficiently numerous to attract the atten- 
tion of Mr. B. S. Ogle to them as he was passing in a train. These 
hybrids can be roughly sorted into two groups, both of which are, 
however, much nearer to L . repens than to L. vulgaris , the influence 
of the latter plant being especially shown by the slightly larger size 
and the much paler (yellowish blue) tint, and by the more conspicuous 
orange palate. In 1892 the gap between the two species was 
practically filled up, but the plant with flowers nearly as large as 
those of L. vulgaris was quite rare, not more than eight or ten plants 
being seen. The influence of L. repens in this was practically limited 
to a very few pale striations on the corolla. The plant had narrower 
leaves than normal L. vulgaris , but the variation in the width of the 
leaves in that species is very considerable. To this I gave the name 
X L. Baxterii in the Report of the Botanical Exchange Club of the 
British Isles, 1893, p. 421. The commoner hybrid was one with nearly 
intermediate characters, but which was, on the whole, nearer to 
Z. repens. This occurred by hundreds. 
The year 1893 marked the maximum quantity of individual hybrids, 
when they considerably outnumbered the parent species. The chain 
of intermediates was again practically complete. Since that year the 
hybrids and Z. repe?is have gradually diminished; this is owing to 
the competition of ranker weeds, which is assisted by the gradual 
covering up of the chalk with soil and refuse from the neighbourhood. 
It is probable that hybridization is assisted by the passing trains which 
