MR. D. OLIVER ON THE LORANTHACE EF. 91 
taining a subcoriaceous putamen.” Loranthus is figured by 
Blume, Griffith, and others as with abundant albumen; and so I 
find it in the seeds of the Indian species which I have examined. - 
In L. europeus, too, it is abundant. I cannot doubt, therefore, 
that Mr. Miers has had old or decayed fruits for examination, in 
which the albumen had shrivelled up from the enclosed embryo, 
and that the albumen has been taken for a layer of the pericarp 
(putamen). 
In the notes and technical descriptions which follow, I speak of 
the calyx and corolla of Loranthus, and the perianth of Viscum and 
the genera allied to it. I am fully aware of the reasonable objec- ` 
tions to this application of terms, but I apprehend they are as 
little likely to be misunderstood as any I could employ for the 
purposes of the present communication. 
LongawTHUS, L. . 
I have revised the Sections of this large genus as established by 
Blume, DeCandolle, and Martius, though without a result so satis- 
factory as I could wish. The divisions of these botanists must 
furnish the basis of any distribution of the species into natural 
groups, although the sections of Von Martius appear to have been 
drawn up with reference too exclusively to Brazilian forms, those 
of Blume to Asiatic and Archipelago species. I feel convinced 
that generic value cannot be attached to any of these sections ; 
the principal structural characters upon which they rest applying 
chiefly to the cohesion of the petals, the number of bracts to each 
flower, whether one or three, and the form of inflorescence. The 
more important character afforded by the mode of attachment of 
the anther to the filament, and which, like the Atlantic, separates 
the species of the New from those of the Old World, though not 
without exceptions on both sides, does not appear of such great 
absolute value in itself as to serve as a mark of generic distinction. 
The basifixed or versatile anthers I do not find associated gene- 
rally with any particular set of minor characteristics serving to 
endorse a generic validity. Some forms differing in this particular, 
are in other respects nearly the same. Neither do I believe the form 
of the anthers available for the grouping of species éxcept in a 
subordinate way, though in two or three sections they are certainly 
very different from the rest. The genus, as Loranthus, is a good 
and natural one—well-defined and easily recognized ; but if broken 
up, it will be found, as in other too familiar instances, that though 
many of our genera are recognizable by artificial characters, and 
u2 
