TRANSACTIONS OF SECTION D. DBPT. ZOOLOGY AND BOTANY. 565 



dichogamy, it is clear that, if we still keep to the indefinite mode of flowering, the 

 lower or older flowers on any given plant will he in the second or male stage hy 

 the time the younger upper ones have reached the first or female stage, seeing that 

 in this case the stigma is developed before the anthers of each flower. Scrophidaria 

 nodosa affords a good example of a plant having proterogynous dichogamy, and to 

 it the present paper chiefly refers. In this plant the stigma, after fertilisation, is 

 removed out of the pathway to the nectar by the bending back of the style on the 

 outside of the corolla ; thereafter the stamens straighten out to occupy the place 

 formerly held by the stigma, and begin to shed their pollen in this position. The 

 corolla of this flower, as usual in proterogynous plants, is small and obscurely coloured. 

 There is also a lax inflorescence, the flowers not being crowded together on one side 

 as in proterandrous and highly coloured flowers, but scattered all round, so that as 

 a whole the plant is not easily discerned from a distance, nor does it readily strike 

 the eye as a conspicuous object. That it is truly dependent on insects and not on 

 the wind for the transference of its pollen, is proved by the presence of a well- 

 developed nectariferous gland, and by its emitting odour. Its inflorescence so far 

 agrees with the indefinite form that as a whole the older flowers (in this case male) 

 occur lower down than the younger (female) ones. Among such inconspicuously 

 flowered plants proterogynous dichogamy seems to prevail, just as the proterandrous 

 is characteristic of highly coloured flowers. Hitherto it has not been shown in 

 what way an entomophilous or insect-fertilised plant could possibly profit by a 

 small uncoloured corolla, nor has any reason been given why this apparent dis- 

 advantage shoidd be generally associated with proterogynous dichogamy. The 

 procedure of a wasp visiting Scrophidaria nodosa afforded the solution of this pro- 

 blem. The manner in which it proceeded was quite exceptional. The first flower 

 on the stem which it visited was the top one ; from this it passed to the others in 

 a somewhat irregular manner going downwards, and finally left the plant from the 

 lowest flower. The same thing having been observed repeatedly, a key to the whole 

 question was furnished ; for any one who has watched bees collecting honey from 

 flowers must have observed that the bee goes to the bottom flowers first, and then 

 visits those next above on the same stalk in regular succession from below upwards. 

 Now, that this order of visitation is of importance in reference to the cross-fertili- 

 sation of a plant, will be obvious if we bear in mind a fact which the experiments 

 of Mr. Darwin (' Cross and Self-fertilisation,' p. 299) have clearly demonstrated, viz., 

 that " a cross between the flowers on the same plant does not at all increase the 

 number of seeds, or only occasionally, and to a slight degree." Now, were an insect 

 invariably to visit first those flowers on a plant which are shedding their pollen, 

 and then to pass to those with mature stigmas, clearly the usual result of this would 

 be that it simply removed pollen from the anthers of the flowers in the male stage, 

 and deposited it on the stigmas of those flowers on the same plant which happened 

 to be in the female stage. The effect of this would be little or no better than self- 

 fertilisation. This is well illustrated in the case of Gladiolus. In this strikingly 

 beautiful plant we have well-marked protandry in association with a highly con- 

 spicuous perianth, while the development of the flowers is from below up, so that 

 the older lower flowers have shed their pollen by the time that of the upper ones 

 are ready. The anthers are at first bent forwards and downwards, so that they rub 

 their polien on the back of a bee entering the flower ; meanwhile the immature 

 stigma is above and behind them, quite out of the way of an insect going into the 

 flower. When the pollen is shed, however, the stamens straighten up out of the 

 way, while the stigma lobes expand and bend downwards and forwards, so that they 

 stand in the fairway to the nectar ready to scrape off pollen from any insect as it 

 passes down the tube formed by the perianth. In this case a bee entering one of 

 the lower and older flowers has any pollen that it brings with it from another plant 

 removed from its back by the pendant stigmas of these flowers, and when it ascends 

 to the younger male flowers towards the top of the spike receives a plentiful supply 

 of pollen previous to leaving for another plant. This pollenit will not fail to deposit 

 on the stigmas of the female flowers which it first enters. From this it appears, 

 then, that the whole elaborate arrangements of Gladiolus to ensure cross-fertilisa- 

 tion would be frustrated, were the insects frequenting it to begin at the top flowers 

 and proceed downwards, instead of visiting them in the reverse order from below 



