t 



APPENDIX. 55 



« 



rays, eacli ray formed of three nearly parallel nervurcs ; it is at 

 least three times the length of the calyx. M. Dunal states 

 that Cacahus has the habit and the corolla of Atropa. On this 

 point he appears to me clearly under a mistake, for its habit 

 is certainly more that of a Nolana, being a prostrate herba- 

 ceous plant, with a fleshy angular, stem, and its corolla, as above 

 shown, bears no resemblance to that of Atropa. One of the 

 peculiar features which I have pointed out in this genus, is the 

 remarkable fleshy epigyuous gland, seen on the summit of the 

 ovarium, like the same feature seen in Thinogeton, to which ge- 

 nus it closely approaches in its general habit, and in the form 

 and colour of its flowers: there is indubitably much analogy 

 in this peculiar feature, observable in both these genera, with 

 the still more conspicuous fleshy enlargement of the summit of 

 the ovarium in Hyoscyamus : there is nothing approaching to 

 this structure in Physalis. It is for these reasons that I pre- 

 ferred placing Cacahus among the Hyoscyamea next to Thino- 

 geton, although I have not discovered that its fruit possesses an 

 opercular dehiscence, nor been able to ascertain the sestivation of 

 its corolla. M. Dunal does not appear to have been aware of 

 these facts, but Prof. A. DeCandolle in a note of the Appendix 

 to the ' Prodi-omus ' (p. 690) adheres to the views of that bota- 

 nist on this subject, and reverses the conclusions to which I ar- 

 rived, without attempting to subvert the facts above-mentioned, 

 or annul the reasonings founded on them : he quotes the character 

 I published of Cacahus nolanoides under the name of Physalis 

 nolanoides. These facts remain submitted to the judgement of 

 botanists, and it appears to me that any one who will carefully 

 compare the analysis given of that plant in plate 49 of these 

 ' Illustrations ' with any species of Physalis, will admit that it 

 cannot possibly belong to the latter genus, and that Cacahus is 

 justly entitled to claim a generic distinction. Physalis, indeed, 

 possesses such well-marked features, that it seems a pity to mar 

 its simple and prominent characteristics by combining it with a 

 group so essentially distinct as Cacahus. 



The ample generic character of Witheringia, as defined by 

 M. Dunal (p. 403), and the details he has given from an exami- 

 nation of good specimens of L'Heritier's typical species, W. sola- 

 nacea, confirm the opinion I long ago expressed in regard to 

 this genus : these details, if carefully compared, will be seen to 

 differ in no respect from the characters presented by most spe- 

 cies of Saracha of the ' Flora Peruviana.' The reasons for this 

 conviction were given nearly four years since {huj. vol.jp. 1 et 21), 

 when I considered the typical plant above-mentioned as a 

 species of Saracha : to this inference I was led by its striking 

 resemblance to another species closely allied to it, which I have 



