60 BALFOUR—NEW SPECIES OF PRIMULA. 
annulus of P. Wardii, Balf. fil. P. sibirica, Jacq. is said to be 
exannulate. But I do not assent to this statement. I am not 
yet prepared to deal with P. sibirica, Jacq. as an aggregate type, 
because I have not yet had opportunity of seeing the Petrograd 
collection in which it must be particularly well represented. 
But I have dissected the flower of the true P. sibivica, Jacq. in 
such representative specimens as a Dahurian specimen collected 
by Losnin and in No. 883 of Roskevitz presented to Edinburgh 
by the Petrograd Herbarium, as well as in Karo’s No. 54 from 
Dahuria (one of the types of P. szbirica, Jacq. var. brevicalyx, 
Trautv. cited by Pax). In allofthem I find an excellent annulus. 
I also find the annulus in European specimens—in that, for 
instance, collected by Hojman and included in Baenitz Herb. 
Europ., and also in Magnier’s No. 2558 from Uleaborg, which 
Pax cites as his var. integrifolia, Pax of P. sibirica, Jacq. In 
Sir Joseph Hooker’s plant from West Tibet, now in Kew 
Herbarium, the annulus is beautifully developed, although 
the Flora of Brit. India, iii (1882), 487, says the corolla is “ not 
annulate.’’ All this shows that a critical study of the micro- 
forms of the aggregate P. sibirica, Jacq. has yet to be 
undertaken. 
P. Wardit, Balf. fil. is the West Chinese form of the siaohetievs’ 
and is altogether a different plant from the Himalayan P. involu- 
crata, Wall. These two in turn are not the same as the P. sibirica, 
Jacq., of the Flora of Brit. Ind., whatever that may be—it is not 
the North Siberian form of the species. True P. stbirica, Jacq. 
does not occur in the Himalayas, and the Himalayan plants so 
named will have to be described under a definite name. 
I may add also that a number of Tibetan plants have been 
wrongly assigned to P. sibirica, Jacq. and another series of 
Tibetan forms has been named in herbaria P. tibetica, Watt, var. 
intermedia—being regarded as a passage between P. tibetica, Watt 
and P. sibirica, Jacq. But P. tibetica, Watt does not run into 
P. sibirica, Jacq. The former is not a dwarf state of the latter. 
P. tibetica, Watt is apparently a widely distributed plant in 
Tibet, and shows remarkable variations in stature, umbels, 
calyx, and corolla. Some of these are so marked in the dried 
specimens that it appears to be probable that more knowledge 
of the living plants and their relations will warrant segregation 
of forms within the type. As we know it at the present moment 
P. tibetica, Watt—if phyletically more nearly associated with 
the aggregate P. sibivica, Jacq. than with other species,—is a 
species easily diagnosed from P. szbirica, Jacq. by obvious char- 
acters, of which I have found the following never to fail, whether 
in herbarium specimens or living plants :—As anthesis proceeds 
the petals gradually reflex and become finally adpressed to the 
