72 Design for a Floioer- Garden on Gravel, 



These remarks apply more particularly to the annuals above 

 mentioned, and to nolanas, schizantliuses, clintonias, gilias, 

 and others of the same fragile character. The most advan- 

 tageous manner in which annuals can be used in beds is, planted 

 in a sinarle row as an edoinfj to the borders. The centre of the 

 beds may either be sown or planted to come in as the others go 

 out of bloom, and thus a very respectable show may be kept up; 

 but unquestionably the most effective combination, to secure a 

 regular succession of flowers, is, bulbs and spring-flowering 

 herbaceous and shrubby plants for the centres of the beds, 

 autumn-sown annuals to succeed them round the sides, and half- 

 hardy plants or late-sown annuals, which may be sown or planted 

 after the bulbs have died down, or the other plants been removed 

 to the reserve garden, for a summer display. In this way we have 

 three distinct combinations of colour in the season; the beds are 

 always covered ; and, if half-hardy plants are turned out for the 

 summer show, the row of annuals round the beds serves admirably 

 to fill them, and pi'oduces a show of flowers until those last planted 

 require the room. 1 cannot subscribe to the prevailing custom 

 of planting verbenas and other half-hardy plants among annuals, 

 to succeed them, because it has a very untidy, not to say slovenly, 

 appearance ; and the permanent plants have a miserable appear- 

 ance, from being etiolated, for a long time after the annuals are 

 removed. 



In the grouping of the plants you have lost sight of a very 

 im{)ortant rule laid down by yourself, but where I cannot at this 

 moment remember; viz., that " every bed in a symmetrical 

 flower-garden, except the one which forms the centre, must 

 have a corresponding bed^ resembling it in colour, as well as in 

 form and position." But, instead of this, you will see by refer- 

 ring to the plan, that of the two large beds, 34 and 39, instead 

 of being of the same colour, one is stocked with Nemophila in- 

 signis, blue, and the other with Lobel/a lutea, yellow, well 

 arranged, so far as complementary colour is concerned, but in 

 direct violation of the old distich, which, slightly paraphrased, 

 is one of the best rules we have in the grouping system, viz. : — 



" Group nods at group, each alley has a brother, 

 And half the platform just reflects the other," 



Again, in point of height, the two beds above-mentioned are 

 planted with plants that will not exceed 6 or 9 inches in height, 

 while the two smallest beds, 11 and 12, are filled with clarkias, 

 which will grow to double that height. This is wrong; for I con- 

 tend that, to secure unity of expression, it is quite as necessary 

 that the height of the plant should be in proportion to the size 

 of the beds, as that they should properly harmonise in regard to 

 colour, Indeed, of the two evils, tall plants in small beds, and 



