552 Prof. G. Klebs. Alterations in Development and [Juue 25, 



in principle, to the same dependence on environment as algae and fungi, and, 

 with the object of showing how far it is possible to-day to alter and to control 

 the development of flowering plants, I will now proceed to explain more 

 exactly the experiments which I have carried out during recent years with 

 the plant-genus Sempervivum. 



Sempervivum, the well-known house-leek, lives in our countries on walls 

 or rocks, generally in dry places, and prefers high mountains, such as the 

 Alps. The body of the plant consists of a short thick stem, tightly covered 

 with thick sappy leaves, and is called a rosette. The plants can be reproduced 

 either from seeds, or, what is more convenient for our experiments, by a 

 vegetative process. 



In the spring the rosette develops short small runners which, after growing 

 some time, in turn develop rosettes at their apex. The young rosette, having 

 taken root in the earth and become separated from the mother-plant, repeats 

 in the next year of its life the same process of reproduction. It is therefore 

 easy to propagate the plant, and all my material for the experiments with 

 the species Sempervivum Funkii is derived from a single plant. 



According to the external conditions, a rosette can bloom in the third or 

 fourth year of its life. Then, without any formation of daughter-rosettes, 

 the stem, hitherto quite short, begins to lengthen, and, after attaining 

 a length of from 12 to 16 cm., commences to flower at the apex. Below it 

 from three to five branches bearing flowers shoot out. After the ripening of 

 the fruit, which takes place poorly with Sempervivum Funkii but freely with 

 other species, the whole plant dies at tlie end of the summer. 



The development of the inflorescence requires a long series of internal 

 alterations of the plant. Quite in the early spring the preliminary 

 processes make the rosettes different to those i)roduciiig runners with new 

 rosettes. This particular state of the plant, destined to bloom later but still 

 without real rudiments, I shall term ripe to floioe.r. Hitherto it has not been 

 possible to cause a rosette to flower in the second year of its life, but we can 

 ■ do so with great certainty in the third year by cultivating the plants during 

 the first year in very rich soil, and for the second year in siiiall pots with 

 very little nourishment and in a relatively dry condition. 



The principal question now is whether such a plant as Semiiorvivuni 

 flowers from innate necessity at a fixed time or, like tlic lower plants, can be 

 altered at will in its dovelopnicnt. Our experiments prove that a very 

 essential alteration can Ik; attained by varinus methods. Wo cultivate 

 a plant, ripe to flower, during the montlis of February ami March in a 

 special soil-bed. The heel itself is heated liy ]iipes buried in tlu; (^arth, is 

 very rich in manure ami iiioislurc, and is covered with glass. Under these 



