212 



KNOWLEDGE. 



[September 1, 1894. 



flowers. A decrease which renders the flower uusym- 

 metrical is of much more frequent occurrence, and, in this 

 case, the decrease seems to he in consequence of the 

 irregular development of the perianth tending in some way 

 to limit the space and nutriment available for the more 

 internal parts. The geranium, for example, has a regular 

 five-merous calyx and corolla and ten stamens ; its ally, 

 pelargonium, which is slightly irregular and more highly 

 specialized in relation to insects, has three of the outer 

 stamens imperfect ; Erodium, again, has regular flowers 

 and only five stamens. The loss of symmetry in the 

 staminal whorl is therefore closely connected with the 

 irregularity of the floral envelopes. This connection is 

 well exemplified in the foxglove order, which includes the 

 mullein Yerbascum, with regular five-merous flowers ; 

 also Calceolaria and Veronica, with irregular corollas and 

 the stamens reduced to two. Most of the ScrophulariacefB, 

 however, are unsymmetrical and irregular, ha\-ing a five- 

 merous calyx and corolla with four stamens, but the fifth 

 stamen is often present in a rudimentary form ; thus in the 

 figwort the nectary is a staminode, while in the garden 

 pentstemon the fifth stamen remains as a barren stalk or 

 filament on which no anther ever develops. As in the 

 case of monocotyledons the most probable explanation of 

 these various conditions is that they represent so many 

 stages in the evolution of genera and species from a 

 common ancestral stock. Where flowers are irregular the 

 departure from the typical form is almost invariably in the 

 direction of better adaptation to insects as fertilizing 

 agents, or of specialization for the visits of particular 

 kinds of insects ; the irregularity may in part be due to 

 the action of the insects themselves, but in any case the 

 loss of symmetry arises from the same cause as the irregu- 

 larity. There is reason to think that the cohesion of 

 sepals or petals is also a provision in favour of insect 

 fertilization, and it is noteworthy that this has also been 

 generally accompanied by a reduction of stamens. In 

 flowers like the rose with separate petals numerous 

 stamens occur, but where the petals are united and the 

 corolla gamopetalous seldom more than five are to be 

 foimd. Ii-regularity is uncommon with separate petals, 

 and where it occurs is not accompanied by loss of symmetry, 

 as we see in the irregular but symmetrical flowers of the 

 violet and sweet pea ; it is otherwise among the gamo- 

 petalie, where an irregular shape is much more common, 

 and is usually accompanied by loss of symmetry. Cohesion 

 occurs much less frequently among stamens than in the 

 petaline whorl ; it is seen, however, in the united filaments 

 of the mallow, and in the anthers of Compositfe and 

 CucurbitacefE. Cohesion is most frequent in the gynoecium, 

 the syncarpous pistil being formed of the carpels combined 

 into one body, very possibly as the result of compression. 



Hitherto we have been occupied only with complete flowers 

 in which four different classes of organs are present. There 

 is a marked tendency in some families for one or more of the 

 whorls to disappear ; this is especially the case with the 

 sepals when, as in the rhododendron, the Compositie, and 

 the UmbellifersB, the flowers are produced in crowded 

 clusters, and the protection afforded by the bracts renders 

 sepals unnecessary. It is the corolla, however, that is 

 much more frequently the missing whorl ; the petals of 

 many of the Eanunculaceae are either very small or wanting, 

 but in the anemone, globe-flower, marsh-marigold, peony, 

 &c., this loss is compensated for by the bright colours of 

 the sepals. The substitution of a petaloid calyx for the 

 corolla in these instances appears to show that there is a 

 competition between the two whorls for nourishment ; it 

 would further seem that in them the sepals have over- 

 powered the petals, and while successful in appropriating 



the share of nutriment have also assumed the attractive 

 functions pertaining to the corolla. 



The dicotyledonous sub-class Incompletffi is so named 

 on account of the absence from its flowers of one or both 

 sets of floral envelopes. This sub-class embraces the 

 goose-foot order, the sun-spurges, the docks, sorrels and 

 nettles, the hop, the elm and the catkin-bearing trees, 

 such as birch, alder, willow, poplar, oak and hazel. The 

 bulk of these incomplete flowers are adapted for wind- 

 fertilization. As the earlier gymnosperms had their pollen 

 carried by the wind, it is natural to suppose that this 

 would also be the case with the first angiosperms, but 

 there is some doubt whether any of the original anemo- 

 philous families have survived to the present time. From 

 the fact that the stamens in these incomplete flowers are 

 mostly superposed to the sepals, it has been inferred that 

 a whorl of petals must have been lost ; and since petals 

 may be supposed to have been acquired for the attraction 

 of insects, the conclusion has been arrived at by some that 

 all existing wind-fertilized angiosperms have degenerated 

 from the entomophilous condition. Reversion to wind- 

 fertilization has occurred again and again in many different 

 entomophilous families ; the common road- weed Plantago,our 

 native Clematis and Thalictrum are without doubt examples 

 of such reversion, but the view that all existing anemophilous 

 angiosperms have degenerated from entomophilous ances- 

 tors, though supported by very many facts, should hardly 

 be accepted as yet without some reservation. The mere 

 opposition of the stamens to the sepals does not aflbrd 

 conclusive proof that a flower or its ancestors ever possessed 

 petals, for, as was shown in a previous article, the super- 

 position of whorls is a more primitive arrangement than 

 even the alternate. The geological antiquity of several 

 families of Incompletse points rather to an early divergence 

 in the direction of adaptation to wind-fertilization from a 

 primitive uuspecialized stock, which had not so far become 

 greatly differentiated in relation to the visits of insects. 

 It is at least a striking anomaly that those families to 

 which the oldest monocotyledons and dicotyledons belong 

 should be regarded as having undergone the greatest 

 possible amount of degeneration. The aroids, which 

 include the familiar Nile lily and cuckoo-pint, are, perhaps, 

 the most ancient of known angiosperms, and yet they are 

 the most degraded of all monocotyledons, so much so that 

 the entire inflorescence is only functionally equivalent to a 

 single flower. They are fertihzed by minute flies which 

 are temporarily imprisoned in the enlarged spathe, but it 

 is quite a reasonable supposition that the aroids have 

 acquired this entomophilous character without passing 

 through any intermediate stages of high specialization 

 either in relation to the wind or to insect agency. 



THE WATER-HEN. 



By Hakry F. Witherby. 



THE water-hen or moor-hen (GaUimda chloropus) is 

 very generally distributed throughout the British 

 Isles. It may be numbered amongst our most 

 famihar birds, but owing to its shy nature its 

 habits are diflicult to observe. 

 The word " moor " seems to have once signified a marsh, 

 and the moor-hen being an inhabitant of marshy places 

 thus received its name, but since by "moor" we now 

 understand heathy and more or less dry land, the name 

 water-hen seems to be more appropriate. 



On almost every lake, pond, or stream on the sides of 

 which reeds and rushes grow, there will be found one or 

 more pairs of water-hen, and when frost drives them from 

 the lakes and ponds they resort to running streams and 



