CONDITIONS INFLUENCING PREVALENCE 223 



only justifiable, but feasible to insist on such a measure 

 being carried out. 



As each pastille costs no more than a Lee-Metford cart- 

 ridge, the cost would not be prohibitory, and the annual bill 

 for invaliding the men who have been educated to use the 

 latter is so heav}^ that it would pay well to adopt any 

 measure likely to diminish it ; for the lessening of the 

 annual invaliding roll by one or two names would more 

 than meet the cost. In most parts of India the month of 

 February would be the best time to apply such a measure. 

 With this exception we possess no means likely to be of 

 service in destroying the adult insect, and therefore must in 

 the main confine our efforts to attacking the more vulner- 

 able larvae. In attempting this the species found in a pool, 

 or even their presence or absence, is a matter of little 

 importance, for almost any collection of water is a possible 

 breeding place ; and the rule should be to leave as few as 

 possible available. All small pools are certainly best dealt 

 with by filling them up ; care being of course taken not to 

 obtain the necessary material by excavating a new hollow. 



In most cases sand from the nearest river bed will be 

 the best thing to employ, as it refuses to hold water, and no 

 harm is likely to result from digging it. A large proportion 

 of Anopheles pools are, however, so small that a dozen or so 

 may be done away with with a single cart-load of sand, and 

 as the results would be fairly permanent this would be far 

 more economical than the use of larvicides. For larger 

 collections of water, pending the adoption of more radical 

 measures, the use of the latter is the only feasible plan, and 

 the only question is as to what is the best material to 

 employ. Unfortunately the larvae are singularly resistant 

 to ordinary poisons, and are capable of disporting themselves 

 for hours in such energetic fluids as the Liq. hyd. bichlor. 

 B.P., or in Fowler's solution. 



Celli (C. M., p. 190, et. seq.) gives the results of an 

 extensive series of experiments as to the quantity required 

 to kill the larvae of a large number of agents, which may be 

 consulted by those who desire more detailed information on 

 the subject ; but their practical outcome is that with the 



