THE WATER GARDEN 171 



or distant effect, but personally I should favour such an arrangement 

 as would produce a surprise. In the illustration it may be noted that 

 no large trees overhang the water. This is as it should be, as these 

 are not, by reason of the shade, at all favourable to healthy growth. 

 Caltha polypetala (the new Kingcup) would add much to the colour 

 effect, while both Menyanthes trifoliata (the Bog Bean) and Butomus 

 umbellatus (the Flowering Rush) would, if not so showy, add to the 

 attractiveness in their season. For growing in the water absolutely, 

 Richardia cethiopica is most valuable in such groupings as this. It will 

 continue in flower for a long period in the summer season. 



Methods of Planting. Various methods of planting can be adopted, 

 but the system I have practised from the very first I have found to 

 answer well. To begin at the beginning, in our case here I would 

 state that the first order that was given was for a dozen varieties. That 

 would be about eighteen years ago. The cost of all of these did not 

 amount to ^3 ; but I should state that the plants were all small, 

 though well rooted and healthy. When unpacked on arrival from 

 M. B. Latour-Marliac I placed them securely in shallow Strawberry 

 punnets. These held them quite comfortably, each plant being tied in 

 to prevent its escape by any movement of the soil. These, with the 

 exception of N.flava^ all throve well the first season and survived the 

 following winter, which was a very severe one. The plant of N. flava> 

 however, succumbed during that trying winter. As I did not then feel 

 assured of their hardihood, I covered the ice, soon after it formed, 

 with some straw litter to prevent, if possible, a thick coating over the 

 Lilies. I followed this system of protection for a few seasons after- 

 wards. Finding, or at least deeming it not to be essential, I ceased to 

 cover them, and have never done so since. These little plants of the 

 first season grew so well, that a few flowers were produced the second 

 year from planting. At two years from their receipt I lifted them 

 carefully. The punnets were, as a matter of course, decayed ; but the 

 roots all lifted well. I then transferred the plants to small, but old, 

 nursery rounds of about 2 feet in diameter. 



The best Soil. The soil I used then and I have found nothing 

 to answer better was, first, a layer of decayed leaves over the bottom 

 of the basket; then some broken-up turfy loam and road scrapings. 

 Into this soil the plants were firmly placed, being again tied down 

 to prevent floating. That season the eleven plants grew surprisingly 

 well, and flowered quite freely for their size. 



Dividing the Plants. In about three years afterwards I found it 

 necessary to divide the stronger-growing varieties. These were N. 

 Marliacea albida, N. M. chromatella, N. M. rosea, and IV. M. carnea, all 

 of which are well known as vigorous growers, even under adverse cir- 

 cumstances. I then used larger nursery rounds of from 3 feet to 4 feet 

 in diameter. It was somewhat of a difficulty to lift the plants that 

 time, so well had they rooted into the muddy bottom of the lake. 

 They were slid back into the water ; then, by means of a cord, the 

 baskets were pulled into greater depth from the other side. Divisions 



