4 PEELIMINART EEMAEKS. 



matioix is negatived by every other point of structure. There can be no doubt, however, 

 that fertilization in the Orchidacece is really accomplished by the agency of insects. The 

 correctness of the foregoing r^sum^ as to the structure and relative positions of the 

 sexual parts of the flower in the Asclepiadacece, is well confirmed by the admirable 

 analytical figures drawn by Rio Creux, under the auspices of Prof. De CandoUe, in 

 thirty-seven genera of the family : here the clavuncle is constant, generally peltate and 

 pehtagona], sometimes club-shaped or subcylindrical ; but in all cases the five angles of 

 that process are glandiferous, exuding a peculiar juice : the development during growth 

 appears to me to show that self-fertilization foUows as a necessary consequence, without 

 the aid of external agency. 



On the other hand, we cannot pass over the different view entertained by the cele- 

 brated Robert Brown, who states unequivocally that, in all the Asclepiadacece " that 

 have been hitherto examined, the absolute necessity for the assistance of insects is mani- 

 fest " * ; and in proof of this he " considered the evidence complete." 



But neither in this memoir, nor in any other work of that celebrated botanist, as far 

 as I can remember, is any evidence produced to show how this agency is employed, or 

 why it is necessary ; on the contrary, the phenomena so minutely described by him, 

 showing how the poUen tubes are generated, how. they force their way through zigzag 

 obstacles {l. c. p. 725) till they reach the stigma, whence they pass down the stigmatic 

 channels of the style (" mucous tubes ") until they reach the ovules, are all effected 

 through self-agency and natural development, without any external influence. 



An explanation may be offered in regard to the coma of the seeds. This term is here 

 confined to that process at the apex of the seed, formed by the strophiolar enlarge- 

 ment of the testa around the micropyle, which bears a number of long erect hairs in 

 several series : this peculiar process is found only in the true JSchitece and in the 

 Asclepiadacece. The term coma is excluded from those instances where it has been 

 employed to denote the excurrent mass of hairs which emanate from the surface of 

 the testa. 



Returning to our subject, I may here remark that the numerous observations made 

 by me in many former years, the careful analyses obtained in Brazil and since my return, 

 the, plants I collected there, still preserved, those in the herbarium of the British 

 Museum, including many original types, the care taken in the analyses of these, 

 have afforded useful results, novel in great measure, all here interpolated with the 

 recorded descriptions of botanists, and all arranged in consecutive order. I have also 

 been able to identify with specimens in our herbaria the thirty species figured by 

 Velloz under Taberncemontana and JEchites. The evidence thus obtained, copious 

 as it is, is not sufficient to form a complete monograph of all the South-American 

 species. To my extreme regret, illness and other causes have prevented my examination 

 of the rich materials in the herbarium of Kew ; I have not even attempted to examine 

 the plants of this family belonging to the Old "World ; so that much more is required to 

 be done, to form a complete arrangement. 



* Linn. Trans, xvi. p. 732. 



