298 THE GLADIOLUS. 



being no distinctive ditference between them to constitute what is gene- 

 rally accepted as an obstacle to the production of crosses ; consequently 

 it must be attributable to some peculiarity we are not acquainted with. 

 Therefore I argue the necessity of looking to the habitat and locality 

 of each plant and the probable influence it would have on them. Lastly, 

 I would suggest tlie experiment — which, by the by, would be rather 

 tedious — of taking, say three varieties of each kind — the more distinct 

 the varieties, and the nearer the alliance, tiie better — and seeding them 

 together, crossing the ofi'spring of tiie six varieties with each other. 

 Thus a series of years would be occupied before the feature would be 

 obtained, but I think that tlie most feasible way of setting about to 

 obtain it; and I believe it would be brought about by gradually im- 

 parting the constitution of the opposite parent to each other's off- 

 spring. 



" The cultivation of this tribe is very simple. They require a good 

 friable loam, with an admixture of leaf mould, placing a layer of cow 

 manure at the bottom of the bed. The natalensis varieties should be 

 planted in March, about four inches deep, and nine inches apart in and 

 between tlie rows. Tiie early flowering kinds are best protected in pots 

 in a cold frame during the winter months, and planted out in tlie spring ; 

 for their growing so early as November renders them liable to suffer 

 injury from tlie severity of the winter if they are planted out at that 

 season ; hence it is advisable to pot them in the autumn, and store them 

 in a cold frame, merely protecting them from frost. 



" I would strongly recommend grouping these plants, and in doing 

 this, Avould recommend the late flowering kinds to be planted in the 

 following way: — Gandavensis splendens in the centre ; a row of flori- 

 bundus around it ; followed by the orange variety of gandavensis poly- 

 stachius, and Colvillii superba ; filling up the margin, two or three 

 rows deep, with ramosus, this being the most dwarf variety of this di- 

 vision of the family. This arrangement would be productive of contrast 

 of colour, and they would be graduated in height. The early flowering 

 kinds would have a beautiful effect planted in the following way : — 

 Insignis central ; La Princess around ic ; followed by formosissimus, 

 princeps, and Rex rubrorum ; completing the bed, two or three rows 

 deep, with Virgin Queen. The bloom of Gladioli here this season has 

 been truly grand ; and I am sure that if they were to be planted in 

 masses, instead of being isolated, as is too frequently the case, they 

 would be much more highly esteemed. The fact of their being expen- 

 sive may prevent many persons from growing them ; but this, I doubt 

 not, will be soon overruled when they are brought into general favour. 



" In raising seedlings in an ordinary May, I should recommend sowing 

 the seed early in the spring, in a cold frame, say about March, pro- 

 tecting the frame from frost and keeping it shut up for a fortnight or 

 three weeks; in the meantime withhold water, there being sufficient 

 humidity in the soil at that season to advance the early stages of the 

 development of the seed ; this caution is of importance at this critical 

 period, after which, however, water may be given as circtmistances may 

 require. The frame must be kept close, giving air by degrees as the 

 season advances, and stimulating them with heat as much as possible. 



