March 26, 1885] 



NA TURE 



487 



ON PETALODY OF THE OVULES AND OTHER 

 CHANGES IN A DOUBLE-FLOWERED FOkM 

 OF "D/ANELLA CASRULEA" 



A SPECIMEN, kindly forwarded me by Baron Sir 

 Ferd. von Mueller, of a double-flowered Dianella 

 carulea, has several points of interest. It is an addition 

 to the scanty list of double-flowered plants from the 

 southern hemisphere ; it is of interest as having suggested 

 to Robert Brown the establishment of a new species, as 

 was kindly indicated to me by Mr. Baker, while the 

 structural peculiarities it presents are specially worthy of 

 note. With regard to the first point, subsequent expe- 

 rience has shown that the late Dr. Seemann's assertion 

 that there was not ;I a single double-flowered species 

 known from the southern hemisphere," except Ruins rosi- 

 folius, no longer holds good, and, indeed, the number of 

 specimens that have from time to time been forwarded to 

 me by Sir Ferdinand von Mueller from various parts of 

 Australia, leads me to believe that such variations are as 

 common in wild Australian plants as in wild European 

 ones, and that, if there be any defect in this particular, 

 it is more apparent than real, and arises partly from in- 

 sufficient observations, and partly from the relatively 

 smaller number of cultivated plants in Australia. One 

 such instance, that of Tet>-atheca cilia/a, presented such 

 features of interest that I made it the subject of a note 

 in your columns, December 7, 1882. 



Robert Brown's Dianella congesta (R. Br. Prod., 2S0) 

 is described by Mr. Baker in his systematic summary of 

 the Asparagacea? (Jcurn. Linn. Soc.,\w., 1S74, p. 576) 

 as having the flowers arranged in dense tufts, in which it 

 differs widely from all the other species of the genus. 

 Mr. Baker expressly says that he had only seen immature 

 flowers. In the " Flora Australiensis," vol. vii., 187S, p. 16, 

 Mr. llentham alludes to the plant in the following terms: — 

 "Dianella congesta . . . appears to me to be a form of 

 D. carulea with dense sessile cymes; the inflorescence, 

 however, in the specimen preserved is scarcely developed, 

 and almost destroyed by insects." The examples sent by 

 Sir Ferdinand von Mueller are, fortunately, in better con- 

 dition, although, being dried and pressed, they afford 

 little or no opportunity of examining the early stages of 

 development. 



Dianella carulea, as grown in greenhouses in this 

 country, is an elegant perennial plant with grass-like foliage 

 and loose, much-branched cymes of bright blue flowers. 

 Each flower is about half an inch in diameter, and con- 

 sists of a coloured perianth of six oblong, obtuse segments 

 in two rows ; each of the outer segments has five pro- 

 minent convergent ribs, the inner ones have three only. 

 Within the perianth is a row of six stamens, three of which 

 are placed before the three outer, and three before the 

 three inner perianth-segments, from the base of which 

 they are, indeed, not entirely free. These stamens are 

 remarkable for their thick, club-shaped, fleshy filaments, 

 which support a two-lobed anther opening at the top of 

 each lobe by a terminal pore. The ovary consists of 

 three carpels alternating with the inner row of stamens, 

 and are thus opposite to the sepals, and consolidated into 

 a thrcc-locular ovary with axile placentation, and with 

 numerous ovules in each loculus, the horizontally-disposed 

 ovules being arranged in two longitudinal lines. The 

 ovary ripens into a fleshy ovoid or oblong berry of a lovely 

 blue colour, and containing a relatively small number of 

 seeds as compared to the number of ovules. Indeed, 

 according to the published figures there is much vari 

 ation in the number of the ripe seeds, abortion of a 

 large proportion being apparently the rule. 



So much relating to the usual conformation of the flower 

 is necessary for the comprehension of the changes met with 

 in the malformed specimens. The first thing that strikes 

 attention in them is the substitution of masses of flowers 

 densely crowded into glomerules in place ot the light 



loosely branching panicled cyme met with under normal 

 circumstances. These glomerules look like little " Brussels 

 sprouts," but their constituent parts are somewhat fleshy, 

 and rich cobalt blue in colour. It was this crowded con- 

 dition of the flowers that doubtless suggested the name 

 " congesta," applied to this form by Brown. On exami- 

 nation of the individual flowers, many changes are observ- 

 able, and scarcely two flowers present exactly the same 

 characteristics. In most cases a multiplication of the 

 perianth-segments has taken place at the expense of the 

 stamens and carpels, but few or no intermediate forms 

 are met with between petals and stamens, or petals 

 and carpels, neirher are there any indications of stamin- 

 ody of the carpels or the converse. Very frequently 

 the thalamus, or axis of the flower, after having given 

 off several alternating whorls of segments, divides 

 into three or more divisions, each of which, in its 

 turn, gives off successive whorls of densely imbricating 

 blue segments. 



The most interesting changes, however, are to be 

 sought in flowers which have not undergone such a 

 serious amount of perturbation as those above-de- 

 scribed, and of these a few may be found here and 

 there wedged in among their more full-blown companions. 

 Unfortunately the flowers are so densely packed, and the 

 state of the specimen such, that nothing can be learnt as 

 to the relative position on the inflorescence of these less 

 distorted flowers. The perianth in these cases is normal, 

 but the stamens present some significant changes. The 

 thickened fleshy filament is replaced more or less com- 

 pletely by a slender ribbon-like stalk, not, as in the 

 natural state, continuous with the base of the anther (basi- 

 fixed), but attached to the back of the anther, a little 

 above its base (dorsi-fixed). This would seem to be an 

 indication that the thickened portion of the filament in 

 the ordinary flower is really an anther-lobe in a state of 

 arrested development. 



It will be remembered that Clos and also Goebel are 

 of opinion that the anther is a distinct organ, without direct 

 relation to the lamina of the leaf, and the first-named 

 author considers the filament and its continuation the 

 connective, to be the representative of the median nerve 

 of the petal (Clos : " la feuille florale et le filet staminal "). 

 It would occupy too much space to enter into a discussion 

 on this point : suffice it to add that, in addition to the other 

 changes noted, the anthers in these flowers open by 

 longitudinal slits at the sides, and not by pores. The 

 ovary presented different conditions in different flowers. 

 In almost every case it was preternaturally enlarged, in 

 some instances it was converted from a trilocular to a 

 unilocular condition, owing to the edges of the carpels 

 remaining "valvate," and not inflected, the placentation, 

 of course, in such cases, being parietal, not axile. In 

 other flowers the ovary was represented by three sepa- 

 rate, but closed carpels, a retention of juvenile or primord- 

 ial character, and which, probably, may also be taken as 

 an indication of the condition of the carpels in the 

 progenitors of the Liliaceae. 



But these changes in the carpels are of less in- 

 terest (owing to the greater frequency of like mutations 

 in other flowers) than are the appearances presented by 

 the placenta and by the ovules, changes unlike anything 

 that has been observed in Monocotyledons, so far as I am 

 aware. These changes in the placenta in the case of the 

 closed unilocular carpels consisted in the outgrowth from 

 the ventral suture of two narrow, parallel, longitudinal plates 

 of a bright blue colour, extending the whole length of the 

 carpels. In flowers in which this petalodic condition of 

 the placenta was present there were no ovules. Are these 

 petal-like processes to be considered as outgrowths from 

 the ventral suture— i.e. of foliar origin — or are they to be 

 regarded as springing from the thalamus (axial), and con- 

 genially adherent to the edges of the carpel ? Unfor- 

 tunately there is no means of obtaining a definite reply 



