618 : DR. M. T. MASTERS ON THE PASSIFLORACEZÆ. 
quadrangularis*—a fact which he considers proved by the anatomical structure of the 
eoronal threads, as also by the circumstance that if the processes in question be early 
removed the flowers remain scentless. 
In repeating this experiment, however, I have not been able to satisfy myself of the 
absolute correctness of this statement, though indeed it is one which would seem very 
likely to be correct. So far as I have observed, the chief seat of the perfume is in the 
, nectary itself. Professor Morren attributes to the conical, pimple-like cells of the epidermis 
of the eoronal filaments the formation of the odoriferous principle. "These peculiar cells 
are found on the surface of the petals and in the nectariferous portion of the tube of the 
flower. The plant, says the Belgian professor, relieves itself of its excess of carbonaceous 
material by means of these two secretions, the sugary fluid or nectar and the aromatic 
principle. That these peculiar papilliform cells, constituting what was called by Prof. 
Morren conenchyma, are often associated with secreting surfaces, there can be no doubt; 
but as they occur on the surface of petals of flowers destitute of perfume, we must await 
further evidence before we assume that in the Passion-flowers these cells really secrete 
the odorant principle. 
Movements of the Stamens, Styles, and Anthers.—The different positions assumed by 
the reproductive organs have been already referred to. I am not in a position to throw 
any light on the cause of these movements, of which, so far as I have seen, the minute 
anatomy offers no explanation. I can only surmise that the differences in position 
result, as they are known to do in other similar cases, from changes in the contracti- 
lity or turgescence and consequent tension of certain cells as contrasted with the flac- 
cidity of others. But this is a matter requiring much additional investigation. 
The peculiar movement of the anther from introrse to extrorse has been attributed by 
some to gravity merely. The broad filament ends above in a fine thread, which is con- 
tinuous with the anther below the centre of its dorsal surface. It was therefore 
assumed that on the expansion of the flower the superior weight of the upper portion of 
the anther, in conjunction with the fineness of the pivot on which it is poised, was suffi- 
cient to cause the anther to tilt overt. This explanation was combated by Morren, who 
adduced several reasons for rejecting this explanation, and considered that the fine point 
terminating the filament acts as a spring, which, under certain conditions, pulls the 
anther over. This explanation I believe to be correct, and in harmony with the singular 
structure of the organ in question and before alluded to. 
It would be interesting to compare, in a fresh state, the anatomical structure of the 
filament at its point of attachment to the anther in those genera, such as Modecca, 
where the anther is basifixed and introrse. 
Movements of the Tendrils, &c.—Under this head I need only refer to the valuable 
communication of Mr. Darwin, to whose observations 1 have nothing to addi. 
* Dodonæa, part ii. p. 21. 
- "t St.-Hilaire, © Leçons de Botanique, ou Morphologie Végétale,’ p. 439; Morren * Dodonza,’ part ii. p. 23. 
+ Darwin, Journal of Linnean Society, vol. ix. p. 89. 
