SEEDS — nYBEIDISING. 13 



Ko binding is required round the junction of the stock and scion, and if 

 the plants are placed in a. rather shady position for a few days and kept 

 dry a union will soon be effected. When the scion and stock are slender, 

 and if nearly equal in size, they may be grafted by cutting the former at 

 the base into a wedge shape, with a corresponding notch in the latter, 

 tying them carefully round in this case, but not too tightly, or the surface 

 will be injured, and probably decay. Nearly all the species in the different 

 genera of the Cactus order unite readily with each other, but it is not well 

 to select a plant for a stock which differs greatly in quickness of growth 

 from the scion or vice versa, for the success of the union is very uncertain 

 in such cases. 



Seeds, — Though the majority can be easily raised from seeds, this 

 method is rarely resorted to except to obtain new varieties or hybrids. 

 The seeds are mostly small, and should be sown in a compost of three 

 parts sand and one of loam in a temperature of 60° to 65°, The best time 

 is immediately the seeds are ripe, otherwise the spring months are the 

 most suitable. Germination takes plac&in from one to three weeks, and 

 the progress of the seedlings is often very slow for some months, during 

 which time the greatest care must be exercised to prevent injury by too 

 much moisture either in the soil or the atmosphere. 



Hybridising. — Nearly all Cactaceous plants produce abundance of 

 pollen, and are readily crossed, especially Cereus, Phyllocactus, and 

 Epiphyllum, to which the efforts of hybridisers have been principally 

 confined. By judicious crossing some handsome forms have been raised, 

 and there is undoubtedly plenty of room for further experiments in this 

 direction. For instance, the Mamillarias vary considerably from seed, 

 yet very little has been done in this genus on any systematic plan. In 

 the majority of species the pollen and stigmas are mature at the same 

 time, and therefore if self -fertilisation is not desired the stamens of the 

 flower selected for the seed-parent must be removed before the anthers 

 burst ; and if there are other flowers open at the same time a small piece 

 of gauze or muslin may be tied over it both before and after the pollen is 

 applied. The species of Cereus hybridise readily, and it is only necessary 

 to observe closely the time when the flowers open, as some last but a few 

 hours, often expanding late in the evening. 



CLASSIFICATION. 



A great number of methods of classifying the members of this 

 family have been proposed, but that adopted by Hooker and Bentham 

 in their "Genera Plantarum " is the simplest and best. Two tribes are 

 ormed. The first, termed BchinooacteEC, is characterised by the calyx 



