PRUNING THE BANKSIAN ROSE, 237 



decided, wliile tlie fruit is immature, the plant exactly corresponds 

 •with the figure of R. tripartitus in the Atlas of Cosson and Germain, 

 making one important exception in the total absence of submersed 

 leaves with capillary segments. Fresh examples of R. Lenormandi 

 and R. hederaceus were shown at the same time, in order to illustrate 

 the difl^erences. Two forms of R. aquatilis were also exhibited ; one with 

 the usual large flowers, the other with flowers about half the size, tlie 

 latter being possibly R.; Petiveri. All tiie five forms were collected 

 the day before on Esher Conunon, Surrey. 



Mr. Watson also exliibited examples of the true Viola flaviconns of 

 Smith, in contrast Mith the V. canina of Linnaeus. They were brought 

 to the meeting in order to show by fresh specimens tlie differences 

 between the two (both in leaves and flowers), winch were verj^ per- 

 ceptible. Mr. W. remarked that small specimens of the ordinary 

 V. canina had been erroneously figured in the " Supplement to English 

 Botany," as tiie V. flavicornis of Smith ; and that Mr. Babington per- 

 sisted in repeating the same error in the second edition of his Manual 

 (published that day), altliough he could not fail to know now that tlie 

 application of Smith's name " flavicornis" to the plant of the Supple- 

 ment was an error on tlie part of Mr. Forster. 



Mr. AVatson also exhibited a variety of V. canina with smaller and 

 lighter coloured leaves than usual, and a smaller and pale purple or 

 pinkish corolla. He had found the original root several years ago in 

 Surrey, and removed it into his garden, where it had flowered and 

 seeded year after year ; all the seedling plants springing uj) around it 

 still resembling the parent in size and colour, witliout a single example 

 of tlie ordinary V. canina appearing among them. He adduced this 

 fact in order to show the hereditary repetition of a variety, without 

 instances of reversion to the normal characters of the species, and that, 

 too, in a character usually deemed so inconstant. Such a repetition is 

 frequently accepted for evidence of specific distinctness, and yet cases 

 like this prove its insufficiency. 



PRUNING THE BANKSIAN ROSE. 



This Rose differs widely in appearance from other Roses, and the 

 difficulty experienced by many in inducing it to grow and flower freely, 

 points out the error of treating it as other Roses. To bloom this Rose, 

 do not prune it at set seasons, as with other Roses. It is disposed to form 

 strong shoots in the summer time. Watch for the appearance of these, 

 and so soon as they are about a foot long, pinch off their tops. In 

 consequence of this check they will form laterals, which become ■well 

 ripened and flower with certainty. It is necessary to cut their tops oflP 

 early in spring, and from this period the plants should be watered all 

 lie growing season. 



