166 THE ORCHID REVIEW. [JuLy-AucGust, 1918. 
dealt with (O.R., xxii. pp. 253-254). Apart from this, it is. seldom 
mistaken, and its characters are too well known to need further definition. 
Orchis maculata, latifolia and incarnata are the three British members 
of the very natural section Dactylorchis, and we believe them to be 
thoroughly distinct, though their natural limits have been obscured by the 
camouflage of hybridity where they grow intermixed. To this we attribute 
the confusion into which the group has fallen. All the three possible 
combinations have been recognised in Britain, O. maculata x latifolia = 0. 
Braunii, O. maculata x incarnata = O. ambigua, and O. latfolia xX 
incarnata, usually called O. Aschersoniana, though the point requires con- 
firmation, as pointed out above. An alternative theory has been advanced, 
namely that they originally formed a single species, and as regards O. 
latifolia and O. maculata that those which established themselves in marshes 
developed in a different way to those which grew in dry ground; that the 
down forms have become fairly stable, while the marsh forms show every 
kind of intermediate. We are not concerned here with the causes that led to 
their separation, but it is significant that it is in the marshes that they have 
come together again, and afforded the opportunity for hybridisation, and 
it is here that the multitudinous intermediates occur that seem to defy all 
attempts to classify them satisfactorily. It is the way with most hybrid 
intermediates. R. A. ROLFE. 
STRUCTURE OF THE OPHRYDE#.—About twenty years ago we called 
attention to a “ novel” view of the structure of the anther of the Ophrydee 
just expressed by the late Sir George King (O.R., vi. pp. 264-266). We 
have just discovered that the same view was suggested over forty yeals 
earlier by the Rev. J. S. Henslow, when figuring an abnormal flower 0! 
Habenaria chlorantha (Journ. Linn. Soc., ii. pp. 104-105, t. |, fig. B). He then 
remarked :—‘‘I venture to put the following questions. May not the twa 
anther-lobes (in this genus and others, as Orchis, Ophrys, &c., allied to it), 
usually regarded as belonging to one, viz., the anterior, stamen of the outet 
whorl, belong rather to two stamens (viz. one to each of the two lateral) 
of the inner whorl, their other lobe in such cases being abortive ? Such ' 
structure would bring these genera into closer affinity with Cypripedium, 
where both lobes of the two lateral stamens of the inner whorl are fertile, 
whilst the anterior stamen of the outer whorl forms the prominent 
staminodium of that genus. If this should prove to be the case, the 
rostellum (where it occurs) would be the representative of the anteriof 
stamen of the outer whorl, and not a process from the stigma, 2 i? 
usually regarded.” Sir George King was evidently unacquainted with this 
paper, but the theory does not fit the facts, as we have already eae! 
7 4\4 
