CULTURE AND USES. 239 



slowly to dry off. In about a month they will almost certainly produce 

 strong tubers, which should be kept in dry sand but not allowed to become 

 absolutely desiccated. In March or April tubers of all of the tender varie- 

 ties should be planted in sand in a warm tank. As each stolon gets its 

 first floating leaf it should be carefully detached from the tuber and potted 

 off, and the tuber replanted for the development of additional stolons. 

 With tender day-blooming species (apocarpous) it is important to obtain 

 two or more plants in the spring. One of these should be kept in a 

 6-inch or 8-inch pot, and dried off about mid-summer to form a tuber for 

 the next season, as the flowering plants usually die. Hardy varieties are 

 propagated from the rhizomes, by cutting and planting out in spring. 



Propagation by seed is also easy, but is not possible for hybrids. The 

 ripening fruits are covered with muslin bags ; the mouth of the bag is tied 

 fast around the peduncle, and the bag should also be tied to a stick for 

 safe keeping. After the fruit has burst and its parts decayed, the bag is 

 taken in and the seeds washed from the debris. Should the pods burst 

 unexpectedly and the seeds be found floating on the pond, they may be 

 dipped up with a fine wire sieve tied on a pole. Seeds of the Castalia 

 group must be kept in water in a cool place until ready for sowing. 

 Other seed may be dried and kept in the usual way. It is advisable to 

 sow Castalia seeds as soon as ripe ; their germination is at best irregular, 

 and it is unusual to get flowering plants the first year. The tender varie- 

 ties may be sown in flat pans, barely covered with sand, and placed in 

 warm tanks in February. The young plants should be pricked off into 

 small pots and kept moving rapidly ; they will produce plants of the 

 largest size the first season. Planting out in the pond must be delayed 

 until all danger of frost, or even of a cold rain storm, is over. The first of 

 June often proves too early in this latitude. N. gigantca is especially- 

 tender. The labeling of seeds and plants in water can only be success- 

 fully done on metal labels. Thin copper or brass on which one writes 

 with a stylus, or with asphalt varnish, makes the only satisfactory and 

 durable label. 



Some hints on hybridizing have already been given, but it may be 

 worth while to recapitulate. No haphazard work pays. It is better to 

 get no seeds than to go to the pains of raising a batch of supposed hybrids 

 and find they are all pure species. The flower to be used for seed should 

 be covered with a fine net (finer than mosquito bar) before it begins to 

 open ; do it on the previous day. As soon as it opens (or before) all of 

 the stamens must be plucked out. For the greatest accuracy the pollen 



