5 
subjects being amongst the worst sufferers. Half hardy or 
tender annuals like Martynia proboscidea flowered well and 
produced fruits, but in many cases the seeds of spring-sown 
annuals failed to germinate. 
Herpaceous BorDERS AND FLOowER BEDS, 
The effect of the long drought was very marked on the herba- 
ceous borders. In spite of frequent watering most of the plants 
were very much stunted in growth and in many cases the foliage 
was scorched and dried. Even such things as Asters (Michaelmas 
daisies), although watered, felt the effects of the continued drought 
and their flowering period was much shorter than usual. 
Kniphofias were also poor for although they dislike damp during 
winter, they enjoy and are always at their best during a wet 
summer. Dahlias as a rule grew much taller than usual and 
generally did not flower with their usual freedom. 
ome summer bedding subjects, such as Heliotropes, Salvia 
splendens, Begonia semperflorens, Verbenas, Phlox Drummondii 
and Zonal pelargoniums, seemed to enjoy the exceptionally hot 
and dry conditions. Of course they were watered, but never 
overhead, so that the foliage did not suffer from the effects of 
brackish water. 
Hardy Annuals generally proved failures more or less, with 
the exception of LEschscholtzia californica and Hunnemannia 
Jumariaefolia which, in common with other Californian plants 
enjoyed the hot and dry conditions. Seeds of such plants as 
wallflowers and sweet williams germinated very slowly and 
irregularly, whilst seeds of daisies and Myosotis failed to 
germinate at all except under glass. Cannas enjoyed the heat 
and have never been so fine before out of doors. 
Roses were already weakened by a severe frost in December, 
1920, which caught them whilst their wood was still soft and 
unripened, and they suffered again from late spring frosts. They 
were thus in poor condition to cope with the abnormal summer 
and in consequence made little or no growth until autumn, when 
heavy dews and cooler conditions came. | 
THE Water SUPPLY FROM THE THAMES. 
As a result of deficient rainfall on the watershed of the 
Thames, for many weeks the quantity of water coming over 
Teddington Lock was negligible, in consequence of which the 
sea-water gradually found its way farther and farther up the 
river. The water used for plants at Kew is nearly all obtained 
from the Thames, being first let into the Lake and pumped 
thence to the filter-beds and reservoir. On the water from the 
Lake and from the Thames being analysed it was found that a 
considerable quantity of saline matter was contained therein. 
Owing no doubt to evaporation, the water in the Lake was 
found to be even more salty than the Thames (see p. 13). The 
