18 



KNOWLEDGE 



[January 1, 1895. 



compound, each ray of the primary umbel terminating not 

 in a single flower, but in a smaller umbel. A curious 

 feature is the almost constant occurrence in the ivy of a 

 single isolated flower springing from a point considerably 

 lower than the umbel. On the specimen drawn (Fig. I., 1), 

 there were two of these detached flowers ; their presence 

 may be accounted for on the supposition that the umbel is 

 derived from the raceme. The latter is a more elongated 

 inflorescence, in which the individual flowers are separated 

 by joints or internodes of the main stem. These isolated 

 flowers of the ivy are possibly relics of this earlier condition ; 

 at the same time it is quite likely that some advantage 

 arises to the plant from the retention of this ancestral 

 character. 



Each little flower has five green petals, which become 

 deflexed soon after expanding ; they are of short duration 

 and fall off after a few days. Alternating with the petals 

 are five free stamens ; the anthers open by a longitudinal 

 chink through which the pollen is discharged, and they are 

 attached to their stalks in the manner shown. (Fig. I., 7 

 and 8.) Both stamens and petals are inserted on the top 

 of the ovary, a position which is described as epigynous, 

 the ovary being inferior. These characters at once fix the 

 systematic portion of the ivy in the Calycifloral division of 

 polypetalous Dicotyledons, in which are included, among 

 others, the orders Biisare;c, Lii/uiiiinnsa, and UmbelUfercB. 

 From the last mentioned the ivy is at once distinguished 

 by its five, sometimes four-celled, ovary ; the fruit of 

 I'liiheUifera. being invariably two-celled. Attached to the 

 inner angle of each of the five cells is a single pendulous 

 seed or ovule, so reflexed that its stalk adheres to one side, 

 forming a ridge or raphe. This inverted or anatropal 

 condition is very usual, and has the advantage of bringing 

 the opening or micropyle close to the placenta to meet the 

 descending pollen-tube, which travels along the placenta. 

 Ivy pollen has the adhesive character common in entomo- 

 philous flowers ; the grams are broadly elliptical and remind 

 one of the pollen of the rock-rose. The style is short, and 

 its extremity forms the stigma ; there is no knob-hke 

 expansion, as so often is the case. The top of the ovary 

 offers, as it does also in the poppy, a convenient stage on 

 which visitors can alight. Nectar, in the form of numerous 

 minute drops, is secreted by the spongy tissue on the lid of 

 the ovary. The yellow colour of this secretmg surface 

 adds to the conspicuousness of the flowers; what attractive- 

 ness they possess is, however, mainly due not to the petals 

 but to the bright yellow anthers, and to the flowers grow- 

 ing together in a crowded inflorescence. 



Notwithstanding its season of flowering, the ivy is evi- 

 dently adapted for insect fertilization. The pollen and stigma 

 are not those of an anemophilous blossom. Wmter flowers 

 are often self-fertilizing, but the exposed honey of the ivy 

 accessible to short-lipped insects indicates, as Delpino first 

 remarked, adaptation to the visits of flies. Miiller mentions 

 having seen the ivy visited by wasps. The stamens ripen 

 before the stigma, rendering self-fertilization unlikely, while 

 the position of the parts is also calculated to prevent it. On 

 a rocky piece of shore where the ivy grows in great profusion 

 I recently had an opportunity of observing its flowers in 

 perfection, and in none of those examined had any of the 

 pollen fallen on the stigmas. I was greatly interested, 

 however, to notice how frequently the flowers were visited 

 by blue-bottle flies. These insects were the only ones, 

 with the exception of gnats, which could be seen in the 

 neighbourhood of the ivy during my visits. From their 

 numbers, and from seeing them constantly at work in the 

 blossoms, I was firmly con\'inced that the ivy is mainly 

 indebted to the blue-bottle for the fertilization of its 

 flowers. Noone would di-eam of any connection between 



ivy and blue-bottles 

 little of the grotesque. 



Ivy flowers have a 

 which, if not disagreeable, 

 partakes of the fcetid or 

 ammoniacal character not 

 uncommon in blossoms 

 adapted to Diptera. 

 Blue-bottles, according 

 to Miiller, fertilize the 

 carrion-flowers of Sta- 

 pelia, and these flies, as 

 Dr.. J. E. Taylor mentions 

 in his interesting work, 

 are often so completely 

 deceived by the putrid 

 odour that they make the 

 mistake of laying their 

 eggs on the petals of Sta- 



the combination savours not a 

 faint though distinct odour. 



Fig 



II. — -Ti'ausTerse Section of 

 Young Ivy Stem. 



pelias in greenhouses. The ivy blossoms, in respect of these 

 visitors, might therefore almost be ranked as carrion-flowers. 



One might easily overlook the five minute teeth between 

 the petals, which are all the ivy possesses in the way of 

 sepals. A better example of rudimentary and functionless 

 organs could hardly be named. Such vestiges or heirlooms 

 are held to prove the descent of the ivy from ancestors pos- 

 sessing both a calyx and a corolla. Sepals are protective 

 organs ; the entire inflorescence of the I'mhelUfera: is pro- 

 tected by the bracts, and the sepals, not being required, are 

 aborted, as in the ivy. The bracts at the base of the ivy 

 umbel are so small that they are useless as organs of 

 protection, and this office is therefore discharged by the 

 corolla. The sepals of the anemone, marsh marigold and 

 other flowers are brightly coloured and assume the office 

 of attraction, which properly pertains to the corolla ; it is 

 much more unusual to find petals, as in the ivy, doing 

 protective duty like sepals. Not only are they of very little 

 service in the way of attracting visitors, but their valvate 

 arrangement in the bud is a further approximation to the 

 sepaline character. 



Green, though the normal colour of foliage leaves, is not 

 very common among petals. As all parts of a flower are 

 modified leaves, it might readily be supposed that a green 

 corolla results from arrested development, and represents 

 the early or primitive condition of flowers ; but it is much 

 more in accordance with our knowledge of flowers to 

 regard green petals as evidence of degeneration. Under 

 cultivation, as is well known, the numerous stamens of the 

 wild rose become transformed into the petals of the garden 

 variety. The frequency of such a change renders it 

 probable that petals were originally derived, not imme- 

 diately from leaves, but from stamens. If, as some suppose, 

 all flowers at first consisted of essential organs alone, and 

 if in those which were much visited by insects the corolla 

 arose from an outer row of stamens becoming sterilized 

 and petaloid for the purpose of attracting visitors, then, as 

 stamens are mostly yellow, the probability is that the first 

 petals would be yellow also. In the water-lily, again, not 

 only do the stamens gradually pass into petals, but the 

 latter shade ofl' into green sepals which must, therefore, 

 represent a later phase in the metamorphosis of a stamen 

 than a petal does. If the floral envelopes have originated 

 from stamens, as their condition in the water-lily would 

 lead us to believe, the green colour even of sepals cannot 

 be primitive, and much less can green be the primitive 

 colour of the corolla. Though somewhat speculative, 

 this theory of the origin of the corolla, which will be 

 found elaborated in Grant AUen's essay on the colours 

 of flowers, has still sufficient warrant to prevent 



