GEOGRAPHICAL DISTRIBUTION.
meridians, -while the elevated land that Mr. "Wallace has suggested as existing ia later
Tertiary times where the northern part of the Bay of Bengal now is,' enabled an
extension of the Himalo-Chinese meridian to take place towards the end of the last
period of stability, or perhaps even at the beginning of tho last period of retreat, into
Southern India. The fact that the South Indian species are so few curiously confirms
the idea that it was not in existence diuing eai-lier times. And equally striking are the
facts that one of the species (P. seylanica) is a member of a highly natural group of
forms the rest of which have their seat in the Indo-Chinese division, while the other
(P. PerroUetii) has its nearest allies in two Kansu and Szetschuan species (P. macrosiplion,
museieola)-, for Mr. Wallace's suggestion is made in order to explain how " a few Malayan
types may have migrated to the Peninsula of India and have been preserved only in
Ceylon and the Nilgherics," and these two species of Pedicularis show how equally
applicable it is in explaining the presence of a few members of an Arctic-Alpine genus.
That zonal distribution was at some time possible between the southern ends of the
European and Siberio-Caucasus meridians is perhaps evidenced by the existence in Asia
Minor of P. eomosa. But the evidence is inconclusive since P. comosa extends far along
the Siberian meridian, and may have thus a polar derivation. Equally inconclusive is
the evidence afforded by the existence of forms so closely allied as P. aeciulis and
P. Ecgeliana or as P. asplenifolia and P. alhiflora in the European Alps and in the Himalayas
respectively. For, as regards the former two, we find an intermediate form in Nortli
China, that is, on tlie Himalayan meridian, but nearer the Circumpolar province; as regards
the latter two, we find an allied form (P. pedicelhta) within the Circumpolar province
310W. The latter fact equally invalidates the evidence from P. rostrata (European) and
P . Nordmanniana (Caucasus).^ It should be noted in passing that the two South Indian
species just dealt with are both examples of species with corollas somewhat highly
specialised in most respects (that of P. seylanica being of Rhyncholophous, that of
P . Pcrrottettii even of Siphonanthcus type), and yet tliat both are bealdess. They thus
appear to afford strong confirmation of the general evidence from morphology that the
appearance of a beak to the corolla was a phenomenon of comparatively recent occurrence—
that beakless corollas are the moi-e archaic.
That zonal distribution once took place between the southern ends of the American
and the Himalo-Chinese meridians appears possible from the fact that the members of
the most highly specialised American groups (P. incurva, attoUens, groenlandica) have their
nearest known ally in a Himalayan species (P. excelsa). Confirmatory of this too is the
extraordinary affinity between the Himalayan P. laclinocjlossa and the American P. Parryi?
iJut it also appears as if the ocean barrier which now prevents such communication had
been formed prior to the appearance of opposite leaves as a character in this genus.
Another fact that strongly corroborates the idea that the genus was not only originally
an alternate-leaved one, but that the character of opposite leaves is late of evolution, is
• ' I'roceedinfft of the Royal Geographical Society, xxi, 619, But there is uo evidence that marine conditions Layo orer
prevailed in northern India to the t-asfc of Agra.
2 The well-kDOwn and extremely extensile affinities that exist between the Ballcan Peninsula and Asia Minor are of course
not affected by, nor in turn do they affect, at least directly, this reasoning wliich, for the whole Arctic-AI pi lie flora, is based
on an hypothesis of "Migration;" that for the flora of the Mediterranean basin depends on quite another hypothesis-that
of comparative " Permanence."
' These facts, as in the case of Ihe South Indian specie a, exactly confirm from the evidenco of an A re tic-Alpine genus
the evidence from tropical or sub-tropical genera. Mr. C. de Candolle has shown this as regards Meliaceee; Jlr. Hemsley iis
regards genera like Abelia, common to Mexico and fo Indo-China; and a recent indication of the same tiling is the discovery
by General Collett of an Osteomeles, hitherto suppoied to be endemic in the Sandwich Islands, in Burma.
ZONAL DISTEIBUTIO>r.
seen in the South Indian species already discussed, which both ^j f ^ ^ ^
although in the Himalaya itself there are no. as many opposite-leaved as there aio
of the genus is almost entirely meridional, examples of the
distribution^of the same species along several meridians are not
that of P vMllata, the most striking instances are those of P. and P. Sceptrum,
w h L bfth occur within the Cir-cumpolar province and both extend along the European
I d he t b e r i a n meridians. P- extends along the Chinese me.d. n also as fa.
the Himalaya, but in a so.ewbat different condition ^.tero.lossaj^ P- Sce^^M
on the other band, is represented in Japan only by a closely aU.ed, but pe. e dy
Ltitiot, species (P. ..^on«.«)- P. also occurs m the European and S t o
m Id ans but is altogether absent from the Hia^alo-Chinese. On the other hand tt s
Z ^ Z t è i in the American meridian by the closely allied, although perfectly d.sttnot,
' Z r t r ' t h e species that have become part of the Intermediate Flora, and that are
thus characterised by a zonal distribution, very numerous. In the Temperate Zone, as it
is usually understood, the most striking examples are P. palusins, euphrastotdc., .yloaüca,
and resi'pimia. P. palustm extends from the British isles throughout Europe, biberia,
Kamtschatka, Alaska, and Rupert's Land to Canada,-throughout the entire temperate
complement of the Circumpolar province in fact. P. sylvaiica extends fmm the Bn isli
islands across Europe only ; its extension to South-Eastern Scandinavia is, as Mr. Blytt
has shewn, an instance of comparatively recent iminigi-atioa.^ P. resupinat^ extends
across the whole of Siberia and North China to Japan. P. euphramides extends from
Kamtschatka through Alaska, Rupert's Land, and Labrador to South Greenland. It is
thus a sub-arctic species rather than a truly temperate one. The comparatively recent
occupation of their temperate habitat by these species is evidenced by tlie fact that
they pass from the Japanese to the American meridian uninfluenced by the ancient
cleavage that the truly Arctic flora indicates as separating these.
Species with this zonal type of distribution are not, however, confined to the
temperate complement of the Circumpolar province. The few species that descend to
comparatively low elevations in the Himalaya exhibit the same phenomenan. Thus
P . carnosa extends at elevations of 5,000—7,000 feet from Kashmir to Khasia, P. rex
at similar elevations from Khasia to Kwei-tschou. The parallelism between these and
P . syloatica and P. resupinata respectively is complete. Nor is this all; for just as we
have seen that P. palustris occupies the fringe of the Circumpolar province in its whole
extent, we find P. gracilis at elevations of 6,000—10,000 feet in every Himalayan
district from Afghanistan to Yunnan.
There are no species in the temperate complement of the present European Alpine
province with a zonal distribution like this. And the absence of any such phenomenon
there corroborates very strikingly the evidence concerning the tract derived from other
sources. We have already supposed that an ancient Mediterranean Ocean prevented an
extension beyond their present limits of the European and Siberio-Caucasus meridians.
Mr. Ball ^ and Mr. Dyer have independently arrived at the conclusion that the shores
' Axel Blytt: Essay on the Immigration of tho Norwegian Flora during alternating raing and dry periods, p. 71
= Hull; Spxcilcgium Flora Maroccana ; Jonr. Limi. Soc. xri, 302.
' Dyer: Plant Distribution as a field for Qeographical Research, p, 16-
. Ror. Bor. QAKD. CALCUTTA, YOL. III.