THE PERUVIAN BARK FAMILY. 239 



1871 I had both B. Humboldtii and B. jasminiflora in bloom. 

 I therefore crossed B. jasminiflora with the pollen of B. Hum- 

 boldtii, and the result was about twenty seedlings, from among 

 which the two following were selected for distribution in 1873 

 viz., B. Humboldtii corymbiflora and B. jasminiflora longi- 

 petala. I also, at the same time, fertilised B. elegans with B. 

 jasminiflora, the result being umbellata carnea, umbellata alba, 

 and candidissima, all of which were good, and were sent out 

 in 1873. B' longiflora flammea, also obtained from this cross, 

 is one of the very brightest of its colour, which is salmon-tinted 

 scarlet ; but occasionally some of its petals will sport to pink, 

 especially if grown in too cool a temperature. From the cross 

 just named was also obtained B. Bridal Wreath, a fine hybrid, 

 having the vigorous branching habit of B. jasminiflora, with 

 finer flowers and much larger trusses ; likewise B. alba odorata, 

 a dwarf, compact kind, with flowers of great substance in the 

 form of very short tubes, and, as the name implies, very odor- 

 ous. The flowers, too, are very persistent, often remaining on 

 the plant till quite dead ; they are of pearly whiteness, resem- 

 bling white marble. The four varieties just described were 

 'sent out' in 1872. In that year B. Davisonii, a beautiful 

 white sport from Hogarth, and exactly like it in growth, was 

 introduced from America. B. Maiden Blush, a soft, rosy, 

 blush sport from B. Davisonii, was obtained in 1873, and sent 

 out the year following. B. bicolor, a seedling from B. flava, 

 crossed with B. elegans, has a habit like that of the latter, and 

 very distinct-looking purple flowers with rosy-pink lobes, slightly 

 tipped with white, and a centre or eye also of that colour ; this 

 variety was sent out in 1874." B. jasminiflora flavescens is a 

 hybrid raised by M. V. Lemoine of Nancy, and sent out in 

 1875. It is a seedling from B. jasminiflora fertilised with 

 pollen of B. flava, and bears clear canary-yellow flowers in 

 umbels like those of the seed parent. 



Bud-variation is of very frequent occurrence in the case of 

 Bouvardias, and, as one would naturally expect, it is rarely 

 observable except in hybrid or cross-bred seedling forms, 

 and then the dissociation of the hybrid or mixed characters 

 is often only partial ; and the result of this is a new 

 variety, different from the hybrid and its two parents. The 

 following lucid remarks on this subject are from the 'Social 

 Science Review' for 1872: "Any plant produced from seed 

 requiring for its development the contact of the pollen-tube 

 with the ovule or germinal vesicle, must be held to have mixed 

 characters, and more markedly so in the case of unisexual 

 flowers, either monoecious or dioecious. From this point of 



