446 



NATURE 



\_April i,, 1878 



and enriched by his own observations and experiments. 

 As their name implies, these flowers never open, and in 

 some cases they have been passed over as abortive bud- 

 conditions of flowers of the normal conspicuous type. 

 Their petals are, of course, superfluous, and are usually 

 completely suppressed, or nearly so, the stamens and 

 pistil are also much reduced in size, but though morpho- 

 logically reduced, are physiologically fully developed, and 

 such flowers are very fertile. In fact, in some instances, 

 as in Viola canina, the production of seed is principally 

 dependent upon them, the ordinary flowers, from want of 

 pollen, or the absence of the visits of bees, rarely producing 

 capsules. 



At first sight the suggestion seems a tempting one, that 

 in these curiously degraded flowers, in which all the 

 laboriously-acquired adaptations for cross-fertilisation are 

 entirely discarded, we have a reversion to a less highly 

 organised ancestral type. And this may still to some 

 extent be true, though Mr. Darwin shows that they " owe 

 their structure primarily to the arrested development of 

 perfect ones." In some cases, as Oliver has shown in 

 Campanula colorata, and Scott in Eranthemum am- 

 biguum, the same plant bears as well as cleistogamic and 

 perfect flowers, intermediate forms between the two. 

 What is, however, still more significant, is that the 

 cleistogamic flowers are themselves sometimes the 

 starting point of structural adaptations, to effect more 

 perfectly the self-fertilisation which ordinary flowers 

 have been so marvellously modified to avoid. Thus, in 

 Spectilaria perfoliata the rudimentary corolla is modified 

 into a perfectly closed tympanum, and in Viola canina 

 the pistil is much modified. Mr. Darwin, however, has 

 shown that cleistogamic flowers do not invalidate the 

 general principle as to the disadvantage in the long run 

 of self-fertilisation. After two years' growth, crossed 

 seedlings of Ononis minutissima beat those produced 

 from cleistogamic flowers in mean height in the ratio of 

 100 to 88. 



It seems that the end really gained by cleistogamic 

 flowers is the production of a large supply of seeds with 

 little expenditure ; the plant does the work more cheaply 

 and makes the numbers pay. It is curious to reflect what, 

 relatively speaking, an enormous expense a plant puts 

 itself to in such a case as Viola in producing in the spring 

 a large number of conspicuous flowers furnished with 

 nectaries and all the complicated apparatus needed to 

 insure cross-fertilisation, with the result, perhaps, of se- 

 curing a very few cross-fertilised capsules. Having made 

 these sacrifices, it proceeds during the summer to insure 

 the production of a sufficient crop of less ^costly seeds by 

 the inconspicuous aid of cleistogamic flowers. 



Mr. Darwin, with characteristic ingenuity, adduces 

 another instance of this balancing of conflicting advan- 

 tages in the effort to secure before all things the perpetua- 

 tion of the race. A seed in the ground — to parody a 

 common proverb— is worth a good many exposed to de- 

 predation above it ; and though dissemination is a gain, 

 secure sowing is no less important. Many cleistogamic 

 plants, therefore, having deliberately given up the advan- 

 tage of cross-fertilisation, give up those attaching to 

 change in the place of growth, and bury their fruits even 

 before they are mature. This is the case with Viola 

 odorata and hirta and Oxalis Acetosella. In other in- 



stances — and Mr. Darwin will pardon the remark that he 

 has scarcely dwelt on the distinction — the buried fruit is 

 the product of subterranean flowers. This is the case 

 with Vandellia sessijlora, Linaria spuria, Vicia amphi- 

 carpos, Lathy rus amphicarpus, and Amphicarpcea^ the 

 three last cases belonging to Leguminosce. The distinc- 

 tion is important because, while flowers produced under 

 such abnormal circumstances as on subterranean branches 

 must be necessarily cleistogamic, it by no means follows 

 that aerial flowers which subsequently bury their fruits 

 should also be cleistogamic, and Mr. Darwin very properly 

 excludes the well-known earth-nut {Arachis Jiypogcea){xons. 

 his list, as, though the ovaries are buried, the flowers are 

 conspicuous. In such cases it is possible that the com- 

 parative humidity of the soil favours the maturation of the 

 capsules, and especially so with small herbaceous plants in 

 dry climates. Mr. Bentham in fact has pointed out in the 

 case of Helianthemtcm that a prostrate habit which brings 

 the capsules in contact with the surface of the ground post- 

 pones their maturity, and so favours the seeds attaining a 

 larger size. Cyclamen (in every species except C. per- 

 siami), by 'the spiral contraction of its peduncle, brings 

 its capsules down to the surface of the soil, though it does 

 not appear to actually bury them, as some authors have 

 supposed to be the case. If this is advantageous we need 

 not wonder that the local amphicarpic races of Lathyrus 

 sativa (of which there seem to be several) found in such 

 dry countries as Portugal on the one hand, and Syria on 

 the other, should acquire the habit of bearing actually 

 subterranean fruit. 



The steps, however, by which such a specialised mode 

 of burying the fruit has "been attained as exists in 

 Arachis, are not easy to follow. Of few plants have 

 the structure and habit been more misunderstood. 

 Descriptive writers, from Rumphius to Endlicher, have 

 represented it as having two kinds of flowers — and 

 as being in fact what Mr. Darwin would call andro- 

 moncecious. It really, however, appears according to 

 the careful examination of Poiteau and Bentham to 

 have only flowers of one kind. These are apparently 

 stalked, but the long stalk is in reality the attenuated 

 calyx tube, which is a very peculiar feature for a legu- 

 minous plant. At the bottom of the calyx tube is the 

 ovary which, after fertilisation, is gradually carried away 

 by the development of a gynophore or subovarian stalk. 

 It is the elongation of this gynophore — and not as Mr. 

 Darwin states, by an oversight, the flower-stems draw- 

 ing the flower beneath the ground — which buries the 

 ovary. The careful observations of Correa de Mello 

 show that though the gynophore may become three to four 

 inches long, the ovary does not enlarge till it is buried, 

 which confirms what has been said above as to the mean- 

 ing of the habit generally. The details of the process by 

 which the gynophore manages to bury the ovary would 

 be a most interesting subject for investigation. 



The obscurity which has attached to Arachis has also 

 extended to Voandzeia, another leguminous plant culti- 

 vated like Arachis in hot countries for its subterranean 

 pods. Mr. Darwin remarks that the perfect flowers are 

 said never to produce fruit (pp. 327 and 341). Correa de 

 Mello, however, never succeeded in detecting the cleisto- 

 gamic flowers, and declares that it is " placed beyond all 

 doubt that the hermaphrodite petaliferous flowers da 



