Harto'g, Organogenic Notes. 341 



cases but little was gained of morphological value, either because 

 of the flower itself presenting no new exceptional features, or 

 because in a few cases the important stages, as so often happens, 

 eluded my research. Out of these studies I published a note „On 

 the Plurality of Axillary Buds in Thunbergia" *), and also an essay „On 

 the Morphology and Affinities of the Sapotacese" **) ; the latter 

 containing also a precis of observations on some allied Orders, 

 which have so far as I am aware quite escaped notice by more 

 recent workers, and which I therefore desire to put more promi- 

 nently on record. Besides these are some others which I kept 

 by me in the hope that, might be able myself to make them more 

 complete, or to offer them to some worker in the same field who 

 could present them otherwise than as mere fragments. To my great 

 regret neither of these alternatives has occurred ; and I now wish 

 to bring forward those notes which refer to points of special 

 morphological interest, or to families whose development is as yet 

 unstudied. I follow the order of Eichler in bis „Blüthen- 

 diagramme". 



ÄsperifoliecB. § Cordiacece. — Cordia sp. (a large flowered 

 S. American species) : it was here perfectly easy to see the origin of 

 each flower from the base of the preceding one; my attention was 

 especially directed to find the alleged dichotomy (as known to me 

 by the reference in Sachs' Botany) , and I could only satisfy 

 myself that the inflorescence was a true scorpioid uniparous cyme. 



Primulinece; Myrsinacece. — In Ardisia solanacea and Ä. 

 crenulata the petals are simultaneous , and arise in front of the 

 intervals between the sepals. After they have become crescentic 

 and almost touch by their inflected edges, the stamens arise as 

 distinct tubercles on the receptacle, one in the concavity of each 

 petal; this indeed being the site of the widest gap, owing to the 

 inflected edges of the petals intruding into the alternipetalous 

 Spaces. The upgrowth of the corolla tube is subsequent to the 

 formation of the stamens. Now in Chrysophyllum , in Ardisia 

 (which is thus shown to agree with Lysimachia, according to Frank), 

 and [m^Primula we find flores hypogynae corolliflorae, 5-merae, 

 staminihus petalis superposita , carpellis alternis , in gynceceum 

 placenta centrali conjunctis. If we follow the dictates of the pure 

 ontogenetic school of morphology and apply the axiom that 

 here as everywhere „allein muss und kann die [rohe] Entwicklungs- 

 geschichte entscheiden", we arrive at the following curious ex- 

 planatiou. 1 ^ In Chrysophyllum the andrcecium is the inner 

 whorl of a diplostemonous andrcecium of which the outer aborts 

 in development. 2 '^ In Lysimachia and Ardisia the stamens arise 

 in the axil of the petals and should represent „pollenbildende 

 Caulomen" sporocauls instead of sporophylls. 3 " In Primula the 

 so-called petals are not really petals (which are absent), but 

 ontgrowths from the stamens simulating petals. This is surely 



") Journal^of the Linnean Society (Botany). Vol. XVII, 

 ") T r i m e^n 's Journal of Botany. March 1878. 



