Vol. XVII. No. 433. 



THE AGRICULTURAL NEWS. 



375 



The outlet should be well above the ground surface (see 

 Le Prince and Orenstein, pp. 130-66).^ 



In soma localities, when ditches become matteJ with algae 

 and other matter, and contain mosquito larvae, it is possible 

 to construct water gates to permit temporary impounding of 

 water, which will enable the ditch squads to flush thoroughly 

 the ditch below the gate, and to remove all mosquito larvae 

 and algae. 



LARVICIDES. 



The ditching, draining, and cleaning of water ways 

 ensure a regular flow, and carry ofi' all surplus water, and 

 thus reduce but do not prevent mosquito breeding. It is 

 necessary to use some additional meaos of control; and for 

 this purpose various larvicides have been applied, but princi- 

 pally kerosene, and crude oils of paratiin and of asphaltum 

 base, as well as creosote oils. 



The question of the eft'ect of oils on )uosquito larvae is 

 most thoroughly discussed by Freeborn and Atsatt, who find 

 that the toxicity of the petroleum oils as mosquito larvicides 

 increases with an increase in volatility, the more volatile oils 

 producing the more marked lethal efl"ect3. The volatile con- 

 stituents of the oils contain the principles that produce the 

 primary lethal efl"ects. The lethal effects are produced by the 

 penetration of the tracheal tissue by the volatile gases of the 

 oils. In the heaviest and least volatile oils, having a boiling 

 point greater than 250°C'., this action may be supplemental, 

 or apparently secondary to the effect of actual contact of the 

 oil with the body tissue, or perhaps to mechanical means 

 such as suffocation or plugging of the tracheae. They found 

 that oils which killed very quickly did so by means of the 

 volatile gases, where^is in case of oils with slow effective- 

 ness, the mechanical suffocation may be the cause of death. 



These authors have set down a number of requirements 

 for a good larvicide: — 



1. It shall have a high toxic power, so that a small 

 quantity may suffice f jr a large volume of water. 



2- It shall kill rapidly, in order that subsequent dilution 

 and weakeriing by rain have as light effect as possi- 

 ble. 



3. It must be uniform in its toxic power, and capable of 



standardization. 



4. It should mix freely with brackish and alkaline 



waters. 



5. It must be harmless to man and domestic animals, 



when in the dilution necessary for larvicidal action. 



6. It shall not be susceptible to rapid deterioration. 



7. It must be inexpensive. 



They did not find any substance which fulfilled all these 

 conditions, but found that a soap, now known as the Panama 

 larvicide, met most of their requireiiienis. 



This was made of the following ingredients: — 

 Resin J 50 to 200 lb. 



Soda 30 ,, 



Carbolic acid (sp. gr. 97) 150 gallons. 



This makes liquid soap which freely emulsifies with fresh 

 water. The carbolic acid must have at least 15 per cent, of 

 phenols, and no greater specific gravity than 0-97. 



This larvicide is manufactured as follows; Heat the 

 carbolic acid in a steel tank with steam coil. When steaming 

 hot, add the resin, and continuously stir the mixture by means 

 of a paddle agitator until complete solution is effected. 

 Dissolve the caustic soda in 6 gallons of water, and add to the 

 mixture. Heat and stir for five minutes. Draw a sample 

 and pour into water. If it emulsifies, the process is complete, 

 and the product may be put into shipping drums which must 

 loe tightly closed. 



(To Id continued.') 



MANURIAL EXPERIMENTS WITH SEA 



ISLAND COTTON IN ST. VINCENT 



IN 1917-18. 



Mr S. C. Harland, Assistant for Cotton Research 

 on the Staff of the Imperial Department of Agriculture 

 for the West Indies, has contributed a paper to the 

 recently issued number of the Fe.s< Indian Bulletin 

 (V"oI. XVII, No. 2), in which he continues his record of 

 manurial experiments with Sea Island cotton in 

 St. Vincent, the first part of which is to be found in the 

 West Indian Bulktm, Vol. XVI, No. S. 



Mr. Harland seems to have reached some impor- 

 tant conclusions as a result of his experiments; the 

 summary of these is reproduced below: — 



The manurial experiments with Sea Island cotton at the 

 St. Vincent Experiment Station have been studied by means 

 of flowering and boiling records for two successive seasons, in 

 which weather conditions were widely different. 



All the manured plots showed increases over the unman- 

 ured. The increase was greatest in the plot to which potash 

 was applied; but a large gain was also shown by the complete 

 manure plots. The remarkable fact was brought out that the 

 addition of phosphate to the potash tended to lower the 

 yield. 



Manures had no effect whatever on the time of maturing 

 of the crop. Manures had no effect on the precentage of 

 bolls to flowers. The yield obtained when two plants per 

 hole were left was considerably greater than when there was 

 only one. The need for further experiments on the spacing 

 of Sea Island cotton is emphasized. Spacing had no effect on 

 the percentage of bolls to flowers. 



It has been shown clearly that, under the conditions 

 under which the experiments were carried out, there is a 

 striking uniformity in the time at which flowering begins, 

 reaches its maximum, and declines to zero. If the sowing 

 date be known, we can state with certainty when most 

 flowers and bolls will be produced, and when the crop will be 

 finished. We cannot predict what the size of the crop will 

 be further than by stating its probable maximum. The 

 size of the crop depends ou weather conditions, and on the 

 incidence of fungoid disease. Thus, in the West Indies, or 

 indeed in any country where cotton depends for its water- 

 supply on rainfall and not on irrigation, the value of its 

 flowering records is very much less than in Egypt. The 

 Egyptian boiling curve is of less amplitude than the flowering 

 curve, chiefly through shedding. The percentage of shedding 

 is said to be practically constant at 40 per cent. In non- 

 irrigated countries the boiling curve represents the bolls left 

 after shedding and disease have taken their toll, and we have 

 seen that the percentage of bolls to flowers fluctuates greatly 

 from season to season. For manurial and spacing experi- 

 ments in the West Indies, it is probably sufficient to take 

 boiling records only, at weekly intervals. I do not think 

 that the practical planter need concern himself with either 

 flowering or boiling curves. For the comparison of different 

 varieties in regard to yielding capacity, both flowering and 

 boiling records are almost essential. It is certain that con- 

 siderable differences exist in Sea Island cotton in respect of 

 the liability to shed bolls. These differences could be brought 

 out adequately, only by a careful study of the flowering and 

 boiling records, 



