366 STUDIES IN SEEDS AND FRUITS 



(2) In the first place, all the ovules begin to respond to the fertilisa- 

 tion of the ovary. This is equally true when it concerns the legume, 

 as in Vicia and Ulex^ and when it concerns the capsule, as in Primula 

 and Iris. 



(3) In many-ovuled flowers failure of a proportion of the ovules is 

 a normal occurrence after the first enlargement due to the stimulus of 

 fertilisation. In the case of plants with capsular fruits, it frequently 

 happens, as in the case of Arenaria^ Stellaria^ Primula^ Sci/Ja, and Iris, 

 that only two-thirds of the original complement of ovules develop into 

 mature seeds. Of the ovules that fail, the greater number abort soon 

 after fertilisation, whilst the remainder proceed a little with their 

 growth and fail as young seeds. The same principle applies to plants 

 with legumes, most of the ovules generally maturing as seeds, whilst 

 of the remainder the majority abort early. 



(4) It is shown that all the ovules conspicuous in the flower can 

 be accounted for in the fruit, the complete ovular complement being 

 made up by the ovules that abort soon after fertilisation, the seeds 

 that fail in an early stage, and the seeds that proceed to maturity. 



(5) It is in the legume that the influence on the form of the fruit 

 of the failures of ovules and young seeds is generally most evident, a 

 marked constriction resulting when the failures are contiguous in the 

 body of the fruit, the degree of constriction being determined by the 

 number of contiguous failures. The same principle applies to capsules, 

 but not usually to the same extent. Yet in fruits like those of Iris 

 and Allium great changes in the shape of the fruit may be thus brought 

 about. In both legumes and capsules, however, but little effect is 

 produced on the fruit's shape by the failure of seeds far advanced in 

 growth. The form of the fruit is determined much earlier in its 

 development. In a word, in the form of the fruit we have the history 

 of the ovule rather than of the seed. 



(6) Dealing especially with beaded legumes, it is first pointed out 

 that abnormal constrictions of pods that are usually symmetrical, as in 

 Faba, Phaseo/us, and Poinciana, are due to the same influences that 

 operate in the moniliform pod, namely, those concerned with the early 

 abortion of ovules and the early failure of seeds. Even woody legumes, 

 such as those of Poinciana regia^ may exhibit constrictions due to the 

 same causes. 



(7) Next come those legumes, as with Albizzia Lebbek and Cytisus 

 Laburnum^ where extensive failures of the ovules and young seeds 

 habitually occur, producing a marked tendency towards the moniliform 

 habit, but not sufficient to justify their being characterised as beaded 

 legumes. 



(8) As examples of the typical moniliform legume those of Erythrina 

 corallodendron and of Sophora tomentosa are taken. Whilst the remains 





