I. Tnheritaiice in Shiiieij Poj^py 69 



(2) Complexitij of lleredifij in Plaids. With plants which are not citlier 

 artificially crossed, or artiticially restricted iroin ciossing, the conception of thc 

 fraternal rolatioii becomes vory complox. If any artificial crossing or restriction 

 in crossing be adopted, \ve can contiric our attention to one special type of brother- 

 plants, but not only is it very difficult to carry out such crossing or restriction in 

 the largc populations which alone seem to give reliable results in heredity, but 

 when carried out we have reached a condition of things very different fiom what 

 happens in wild life, where except for absolutely self-fertilising plants the diversity 

 in brother-plants is itself a factor of the Variation which evolution finds to draw 

 upon. If we examine tlie case of man we find only three types of fraternal 

 rclationship, i.e. whole siblings or half-siblings on mother's or father's side. But 

 in plants great coinplexity is introduced by thf multiplicity of ovaries and pollen 

 sacks, and by the possibility of seif- or cross-fertilisation. If we put on one side 

 (i) the division of the ovary into cells, and (ii) the division of the anther into lobes, 

 we have still (iii) the multiplicity of anthers on the sanie flower and (iv) the 

 multiplicity of tlowers on the same plant to deal with. Are we to consider the 

 plant as an individual because it proceeds from a single seed, and to suppose the 

 rudiments of future seed in all cells of the same ovary and in all ovaries of the 

 same plant of equal valency ? And again is all pollen whether from anthers of the 

 same flower, or from anthers of different flowers of the same plant, equivalent from 

 the Standpoint of parentage ? Or, is the plant to be iooked upon as a colony of 

 individuals, and the flower to be the unit of parentage ? We may even go beyond 

 this and consider tlie individual cell of the ovary or the individual anther as our 

 unit. It is clear that only very wide-reaching direct experiments on artificial 

 fertilisation would enable us to distinguish whether one or other of the many 

 types of plant parentage give offspring more or less alike. Still some Classification 

 of sibship in plant life seems desirable if we are to compare our poppy results with 

 those for insects and animals. 



We confine our attention for tlie time to the flower as the unit of parentage, 

 and denote by a single letter an individual plant; small letters a, a, «"... denote 

 ovaries of different flowers of this plant, large letters ..4, A', A" ... the groups of 

 anthers attached to the corresponding flowers or ovaries a, a, d'.... A pair of 

 plants growing from seeds takeu from the cells of the same ovary will be termed 

 co-ovarial; if they come from two ovaries of the same plant bi-ovarial, if from 

 ovaries on different plants dis-ooarial. If the pair of plants come from the pollen 

 of anthers on the same flowers, they are syii-anthic, if from anthers of different 

 flowers of the same plant di-anthic, and if from pollen of different plants dys-anthic. 

 When fertilisation is from the anthers attached to a given ovary, we shall terra the 

 resulting plant homotropic ; when fertilisation is from the anthers of another flower 

 of the same plant, the result is endotropic, and when the pollen is from a different 

 plant exotropic; the crossings for the same are hoinogumic*, endogarnic and exogainic. 

 Given a pair of plants sprung from the ovaries of the same plant or from pollen of 



* Not to be confiised with the original use of this word for plants which uiature tlie orgaus of botli 

 sexes at the same time. 



8—2 



