2 
ms eS er ~ 
= - 
es eee 
CONSPECTUS TAPULARUM. 19 
ship; for we have separate petals constantly in Ambelia, and often in 
Myrsine; and perigynous insertion in Mesa. But in Heteropyxis the ovary 
is never unilocular, nor is the placentation ‘ free-central ;”” consequently 
Tam advised by my friend Dr. Hooker, and other botanists with whom | 
have consulted, that it cannot with any propricty be placed in Myrsinea. 
Let us next compare it with Lythrariea, to which, as suggested to 
me by Drs. Hooker and Thomson, Heferopyis bears a very considerable 
relationship, greater perhaps than that to J/yrsinee ; insomuch that 
possibly it may yet be proved to be an actual Lythrarieous genus. In 
the first place, it sufficiently agrees with Zythrariee, in the perfectly 
entire leaves, in the absence of stipules, in inflorescence, and in the 
structure of the ovary. Nor do the glandular dots offer any anomaly, 
for such occur, both in calyx and foliage, in some Lythrarice (as in 
Woodfordia (Grislea tomentosa, Auct.)). In the form of the calyx, 
moreover, and very nearly, though not exactly, in its wstivation ; in the 
perigynous insertion of petals and stamens, we have further characters 
of Lythrariee. But in most Lythrariee the stamens are inserted within 
the tube of the calyx, at an appreciable and generally considerable dis- 
tance below the point of insertion of petals, and in cases where they are 
equal in number to the petals, they alternate with them; and even in 
the known diplostemoneous genera of Lythrariea, the outer stamens alter- 
nate with the petals, asin Lythrum. The insertion of the stamens within 
the calyx tube, however, though general, is by no means constant in 
Lythrariee ; for they are placed in the throat of the calyx in Anthery- 
lium, Duabanga, Sonneratia, Punica (if that be Lythrarieous), and in 
Crypteronia. In the last named genus, as Dr. Hooker suggests, we 
have the nearest approach to Heteropyxis ; for had Crypteronia petals, 
they would oppose the stamens, exactly as do those of Heteropyxis. 
Crypteronia, however, differs by its opposite leaves, strictly valvate 
calyx, form of anthers, and unisexual flowers. Nor do we yet know 
whether there is any agreement in structure of fruit or seed. The time 
has not come, therefore, for associating Heteropyxis with Lythrariea, 
although we may admit an affinity, perhaps suflicient to authorize our 
placing it in an appendix to that Order. 
Very few words will suffice to show that the connexion with Rham- 
nee@ is more apparent than real. Heteropyxis indeed agrees with 
Rhamnee in the opposition of the stamens to the petals, and their joint 
insertion at the summit of the disc-lined calyx-tube ; but it differs 
very much in habit, especially in the nervation of the leaves, the ab- 
sence of stipules, and in inflorescence; in the presence of glandular 
dots ; in the estivation of the calyx, and more especially in the poly- 
spermatous placente. Thus the points which separate it from Rhamnee 
are more numerous, and some of them of graver import, than those that 
suggest its relationship to that Order. 
What, then, shall we do with our Heteropyxts ? Until its ripe fruit 
and seeds shall have been seen and examined, its true place in the system 
must remain uncertain. The name, derived from érepos, and zvéis, 
was given to it on the supposition = it might prove to be an anoma- 
c 
