--~

BULLETINS FROM

THE INSTITUTE FOR MEDICAL RESEARCH, FEDERATED MALAY STATES.

No. 2 of 1939.

The Practical Application of Anti-Malarial Measures on Malayan Estates

BY S. C. HOWARD, M.D., CH.B.,

Estate Medical Practitioner.

The Control of Urban Malaria (Kuala Lumpur) BY

w. E. HOLMES, M.A., M.B., CH.B., D.P.H., D.T.M. & H., Health Officer, Kuala Lumpur.

in Malaya

.KU~~A LUMPUR: PRINTBD AT TUS FSDSRATBD MALAY STATKS GOVSRNMBNT PRK8S.

1939.

BULLE'TINS FRO M

THE INSTITUTE FOR MEDICAL RESEARCH. FEDERATED MALAY STATES.

No. 2 of 1939.

The Practical Application of Anti-Malarial Measures on Malayan Estates

BY

S. C. HOWARD. M.D.. CH.B., E state Medical Practitioner.

The Control of Urban Malaria (Kuala Lumpur) BY

w. E. HOLMES, M.A., M.B., CH.B .• D.P.H., D.T.M. & H., Health Officer, Kuala Lumpur.

The Control of Rural Malaria in Malaya BY

T. WILSON, M.B., B.CH., D.P.H., D.T.M . & H., Health Officer. Lower Perak.

Being lectures given a t the Institute for Medical Research, l"edera!;ed Malay States, during the Sixth International Course in Malariology, 1939, al't"anged by the L eague of Nations , E as tern

Bureau, Singapore.

hl ' ALA J" U ~I1' U R : PIUNTK.U AT Tl-llt FIWIWATIU> MAU,y STAl'P.S <";(JVtm~ ~IK:o-.:T l'IU:SS

193!J.

THE PRACTICAL APPLI GATIO N OF ANTI-M ALARIAL MEASU RES ON MALAYAN ESTATES.

B y S. U. H OWARD, M. D., CH.B.,

Estate Medical Practitione'l' . In approaching the problem of the control of malaria on

rllbber estates in this country, fOll r points present themselves for cO l1 sideration-

1. Th e species of malaria vectors . 2 . The nature of their breeding places . 3. The s.y.·tem of anti-malarial organisation. 4. Th e types of anti-J1Jalarial measures.

I.-THE MALARI A VECTOHS. Although Gater, in his book "Aids to t he I dentifi cation of

Anophelin e Imagin es " published in 1935, records 41 species and sub-species of Ano pheles in Malaya, only five of these are well­recogni bed malari a carriers. They are A. mac1l,latu8, A. umbro ­S itS, A . novum b'l'os1t8, A. sundaicll,s and A . ba'l' birostris. Others have from ti me to t ime been incrim inated , but in general t hey have proved of m inor importance.

2.-THEI R BREEDING P LACES. I These fi ve species have each certain distinctive breeding

characteri stics, a knowledge of which is a guide to the apPl'Opriate lil easures for their control.

(a) Ano pheles 'IILa c'ulai1I,s .-This species is known as the ti L1l11i t stream and .' eepage breeder. It is, so fur us Lhe lua jority of rubber estates in hilly land are concern cd , Public E nenlY ~o . 1. Now, although sunli t streams and seepages may ' be regarded as its optim um habitat they form but half the story of th e poss ible breedin g places this most adaptab le malaria ca rrier ca.n make use of. During the past four years much new agricultural activity has taken placc all over t he Pen insula in t he form of replanting, and in spite of extensive and well­organised anti-maJal' ial contJ:'ol of ravine, swamp and seepage areas, t here have been widespreaJ outbreaks of malari a.

It follows that some new factor has been introduc d which favoured the breeding of this malari a vector . This factor is now recognised to be new only in the degree it h a' afforded "maculatus" extended opportuni ty for breeding. It ,,'as long ago known th at " maculatus" could breed in standi ng water if flS ociated wi th recent di sturbance of the earth and exposed to ft few hours sunshine each day . Prior to the exten,:;ive replanting programmes which commenced in 1936, such standing water as was provided by earth wells, s ilt pits , isolation drains, ca t t le hOQ£ marks in wet grass, etc., were all recognise.d as possible " maculatus" breeding places . R eplant ing introduced three new .factors, each of which produced conditions tending to favour t he breeding of "maculatus" to an extent not hitherto antic ipated. Those in charge of anti-malarial organisations on estates were caught un awares and although the location of the breeding places was early defi ned , it took three years to persuade the planting commulJity of the facts an d to obtain co-operation in combating these new sources of potential danger to th e health of th e community. The factors referred to are-

(a) Th e formation of holes following the removal of old rubber tree roots .

\\ -,' 2

(b) The digging of replanting holes which were left open for two or three months to weather the soil.

(c) 'rhe construction or tenaces on hillv land for the purpose cf contour pla nt~ng (a meas-;"re to minimise soil eros ion) .

\iVherever the soi l contain s a preponderance of clay these measures may res ult in holes capable of holding water which, • under the conditions of replanting, is naturall y associated with considerable disturbance of the oil and is exposed to the sun all day long. Under these circumstances "maculat-us" finds many happy breeding grounds .

To impress upon you the serious breeding potentialities arisillg fro111 these agricult ural sources, I will describe what happens on a hypothetical 100 acres of replanting.

The old tree stance of such all are.a may be taken at 70 trees to the acre. When these are felled and the old roots removed, t he 100 acres is left pitted with 7,000 holes about 2 feet in depth and 3 feet wide. In between the rows of these holes the replan ting holes are made, 2 feet square and 2! feet d~ep, to the number 140 holes to the acre , 01' 14,000 holes in the 100 acres . Thus there are 21,000 possible water-holding holes the earth of wbich has recently been t urned, a.nd wbich are exposed to the fu ll blast of tbe s un.

It i. ' obvious that all these holes will not ho ld water, but if, say, t wo pel' cent. of them do so, and only a couple of dozen of these beccme breeding places of A. maculattls, there will thereby be liberated a sufficient army of anophelin e vectors to give ri se to a serious outbreak of malaria . I have known several hundred "maculatus " . larvffi to be collected from a few such old tree holes during a s:ngle survey. To these artificially made breeding places m.ust be added the terraces wh ich are necessary on hilly and undulating land to prevent '.vash. They are from 4 to 6 feet wide, constructed on the contour, and are made to slop backwards. Th ere are often 25 chains of terraces to the acre, i. e., 30 miles to tbe 100 acres . B etween the terraces cover crops are grown which, within a year or so, spread over the teri'acesand cover dmins , holes , contour drains, etc ., making it very diffi cuit to spot standing water, but covering it insufficiently densely to prevent breeding. Th e danger of such me~tsures giving ri se to cOl1 centmted bre.eding places of A. mac11.lat'lls is very obvious. L arval surveys have confirmed the fact flnd the wide­spread outbreaks of malaria that have startled the country sin ce the commencement of repl;:mting in 1936 prove tjhe case . Thus I wou ld repeat that so far as rubber estates, particularly those on hilly la.nd, are concerned , "maculatus " is P ublic Enemy No. 1.

(b) Otli e-/' Vecto1's.-Since to-day I am speak ing not of ant i­malarial measures in general but of l1W O\\'n experiences in the control of malaria on estates in t,he State of Negri Sernbilan, and sin ce A. l.lm b1'OB tlS, A. noVtlmb1'08US, A. 8'U.ndaic1l.s and A. ba1'bil'Ost1'i's have not besmirched the': r family escutcheons by seriously marauding the districts I work in , I shall, so far as they are concerned, limit my remarks to general observations.

Anopheles 11.rnb1'osus and novum/J1'OStlS are shade breeders with a predilection for j,ungle-covered swa.rnps wh ere the sdil is of tL peaty nature snch as so often is the case in the flat coastal a.reas. 1'he shade does not need to be so heavv as the term " jungle-covered" would seem to indicate . Ev~n wh ere the

3

jungle has been. felled an'd replaced by light secondary growth these shade breeders may reappear if the type of earth is suitable and the water lies in stagnant pools or drains.

Anopheles s1tndaiclls is a brackish water addict and is con­sequently only found within tidal areas. It made dark history in the early days of the development of Port Swet tenh am ""here

II Dr., now Sir Malcolm "\'latson gained one of his early ba ttles against malaria. This Anopheles is not known in Negri Sembil::tn. I, therefore, have no experience in dealing with it.

Anopheles barbirostris isa pool breeder which, in the past few years, has gained notorie.ty in Perak anJ Selango1'. It is partial to pools such as over-grown borrow pits and mining holes containing coarse vegetation. I recently found it breeding in a half empty fish pond, the bottom of which wa s covered with coarse grass and aq uatic vegetation.

3.- THE AN'l'I-MALARIAL ORGANISATION. 'rhe system of anti -malarial organisation whieh I practice

is divisible in to three sections-(a) To define the area to be protected. (b) To ascertain the breeding places of anopheline vectors

within the area. (0) To establish an org;:mised system of weekly records

of anti-malaria work done. For the purpose of the practical application of an t i-malarial

measures in this country, it has for many years been agreed that the anopheline vectors concerned normally have an effective flight range that ' does not exceed about half-a-mi le. Un­fortunately, however, observations made in 1937 and 1938 by Dr. W. E. Holmes showed that under the cond ition s introdu ced by replant.ing "maculatus" can covel' about three t imes t.hat distance in sufficient numbers to cause a serious outbreak of malaria. Still, for the purpose of this lecture I shall work on the half-mile limit which suffices under ordinary eircllmstnnces.

(a)-DEFINITION OF 'l'HE AnEA.

'rhe first step in the initi ation of an anti-m abrial orga nisa­tion on rubber estates is to obtain a survey plan showing the disposition of th e ravines and the location of the habitations to be protected. On such a plan a primary circl e is marked having ft radius of half-a-mile, the centre being at the main group of dwellings to be protected, usually the labourers' lines. Secondary segments are then taken from such place.· as European and subordin ate staff bungalows-which are usually situated at a distance of t en 01' twenty chains from. the l.in es and consequently require observation of ravines , etc ., for .a corre.spanding distance beyond th e limit of the primary circle . In this way the area to be organised is defin ed an d the ravines and swamps lyin g therein can be readily traced and marked out in the field .

Each half-mile circle encloses 500 acres which, in t he average hill-land estate, contains four miles of streams, ravines !lnd swampy areas. By the t ime the main stream or ravine has been dra.ined and the. seepages have been collected in narrow intercepting drain s, the four miles of drains will have extended to something in the region of twelve miles . If such a control a.rea encloses broad valleys, many miles of agricultural drains will also have to be supervised arid maintain ed by the anti-malarial organisation .

4

(b).-Tm: L ARVAL SURVEY.

The second step is to make n deta iled surve.v fo r Anopheles larvm of all strenms, seepages, pools , etc ., wi thin t.he define.d areft . Such fI s ur vey, if mnde between March and the beginning of June, will give a reliable. general idea of the mala ri al poten ­t iali ty of the area . I have m entioned the period March / June because this is the most active breeding season of A. maCtLla t 'Ll8, the chief ma laria ca rrier on hilly rubber estates . Th ere is a second period extend ing front October to December when a recrudescence of breeding often occurs, bu t which , probably owing to th e heavier rains of that season , is only inireq uently accompan'ied by much H1 Rlari a ,

Although th e p riods m entioned make what may be regarded as a p icture of cause and effect , namely , t he act i.ve bre.edi ng of " maculatus" and outbreaks of rnalaria , such periods shou ld not be regarded as defin ite lim itation s to the breeding of th is J1 lo$quito. Con sid rable var iations both in the breeding and in the incidence of .malari a m ark the h istory of this country, The factors which favo ur such vari ability are as fo11ows-

(i) Seasonal breed ing h abits . (ii) Meteorological condit ions .

(ii i) Agricul t ural and othcr act i\'i t ies ca usin g d isturbance of the soil.

( i)-SEA SONAL ll RE EDI:\'G.

'l' he underly ing biological fa ctors which 111Rke for sensotl al bi'eeding in a coun try so well rained upon and con sequently full of breeding opport un ities ill st ream s, swamps, eep~l ges and pools a ll the year round, h ave not , so fu r as I kn ow, been completely exp lained . Th e fact , how'v r, rema ins th at April / J un e FInd to a l13s ' degree October / Dec mber , in the foot -hill d istricts, show J'eau lad y recurri ng out breRks of mala r ia .

( ii)-METEOn OLOG ICA L CON DITION s .

Th ere is an annual rainfa ll in t his countr y of over 100 incll es. Although October / December are di ·tinct ly t he wettest 11l0nths, period s of heavy rain m ay occur at any t ime throughout the year . Wh en this occurs, fl at and wa mpy land may be stimulflted in to short -tim e breeding activity . If such wet periods coincide with agricultural works s uch as felling of bl:u/'- a"l' or jungle , the m aking of holes p reparatory to planting, e tc ., it be.hoves t he mala1'iologist to be very watchful if an outb reak of ma.laria is to be avoided.

(c).-THE S UHVEY O rtGAN I SA'l'IO N.

B cause of this vari abil ity in the breed ing of " ma culnbus " it will be obvious tha t an occasional larval sur vey will not give f\ t r ue id ea of th e m alarial potent iali ty of an urea r equiring control. For th is reason I organ ised, many years ago, a system of weekly larval surveys on all the esta tes under m y !lUpervisioll . On ly in t his way can one keep on eself reasonably wcll inform ed of the disposit ion , strength and activity of the enemy . ' T ir e larval smv eyor' thereby beco.mos t,he scout of the organisat ion who,'e duty it is to bring in early infOl'm at ion of impendin g enemy attacks .

Again, if th e hospital is s it uated wi thin the malarious aren, as is usually the case , a r egular early morning exam inat ion of the mosquito nets forms a most useful check on tho efficiency

5

of the anti-ma lal'ial work. In like manner, jf there are Chinese "lines " or lwngsies within the protected area, exam ination of their nets at frequent intervals will serve the same purpose.

In order to give precision to these weekly larval surveys it is necessary to organise a system of recording the results. To this end the ravines should be divided into sections, and lettered aud numbered for reference. Such divisions are of an arbitrary nature. No specific distances are taken, but th e na tural division s due to branching of the streams are made use of. These divisions are first mnde on the survey plan. Then a record of the letters and numbers corresponding to the divisions is made on month ly sheets . Two sheets are used; one to record wh en surveys are made and the results obtained, and th e other to show wh en the sections al'e oiled and the amount of oi l used.

In the ravin e', boards tire erected corresponding to th e divisions marked on the plan , and are let tered and numbered. In this way it is easy for anyone sent to make a check survey to know where he is and to mal{ e an accurate control examin ation.

FUl'ther, from such weekly survey records a month ly summary is compiled. and at th end of the year these res ults are recorded on tI separate sheet. In this manner a picture is gradunlly built up showing those ravines that are regular breeding places, a nd those that breed only occasionally or not at all. Such a t able is of great p l'ac ti cal use in estimating when the oiling m easures should be increased and when economies r.an be safely introduced. This system also illustrates the seasonal breeding waves and the influence of extraneous agri­cultural works associated with replanting upon th e variability of Anopheles breeding. For estate anti-malar.ial con trol. where every possible econonlY which does not conflict with efficiency has to be constantly kept in mind, the regular w ekly larval survey Rystem will be found to be the key to the situation.

4.-ANTI-lVIALARIAL CONTROL MEASURES. Siuee " maculu tu ," is the chief vector on In y estates I shall

speak only regarding the methods used to control this species. 'l'h ere are two ]l'incipal m thoel,', in wide geneml use, which

can be depended upon to give satisfactory result .. . Th ey are­(a ) by I lac ing all visible water in pipes underground, as

in subsoil drainin g, (b) by open earth drains .and the \J e of larvicides . such

as prepared oil mixtures, Paris Green, ek.

(a)-SUBSOIL DRAINAGE.

This me'thod involves too high a capital expenditure to be considered to-day an economically practical means of control suitable for use on estates . It i' the ideal mcthod for urban areas wh ere the popul ation is concentrated and consequentl y the eost per head shows up at a reasonab le fi gure. But on estate,' where the population per acre is small, the cost per head 'for subsoiling :111 ravines in th e 500 acres or mol' of the health eirc le, takes on n different complex ion. Moreover, the cost of a subsoi l scheme is not li mited to the initial cost of con ,truct,ion: the work must be maintained. Th e annual cost of this is reckon ed at 5 per cent. of the co, t of construction. 1! or th is amount alone, without locking up the ini tial capital, open drain­age and oiling met(.f;ureS can be carried out on the average estate, and reliable control thereby ensured . Again, by the open drain method there is no sacr ifice of arable land, in fa,ct land

•

6

is often reclaimed and afterwards planted. Wh ereas, when :1 l'avine is subsoil-drained, plant ing must be. restricted to a dis­tance of 80 feet from the nearest pipes .

(b)-OPEN DRAINS AND OILING.

'1'he construction of open earth drains and the spraying of oil at weekly intervals are the usual combined methods of control of "maculatus " in use on estates. They give eminently satis ­factory results if carefully organised and supervised. They should be supported by the system of weekly larval surveys previously mentioned.

COX S'1'RUC TION OF OPEN DRAINS.

Appropriate drainage is the fi rst and most important measure of at tack on the stronghold of •• maculatus". If correctly designed and carefully constructed , drain age alone will reduce the m alarial potentiality of an area by some 75 pel' cent . Given an a verage type of clay soil , the construction of effective anti-malarial d,rains is by no means a difficult under­taking. '1'he followi.ng t hree principles will be. a useful guide-

(a) Don 't be deceived by surface water. (b) Intercept t he seepages. (c) Make t,he drains narrow and dee.p.

'Where m alaria is hyper-endem ic, an estate medical practi­t ioner finds his t ime occupied more with fi eld work than wit h the dispensing of physic. H e must have a reasonable knowledge of local malarial entomology, an interest in drainage construction, and be able to organise' oiling and other larvicidal measures.

To this end he should hold himseJf responsible for the con­stl'llction and upke.ep of all ravine, seepage and swamp drai.nage work. To act merely as adviser on anti-malarial requirements, and leave the construction to the manager of thE:) est ate, is to invite divergence of vie.ws and variations in methods which too frequen tly result in disappointment. This is so because mana­gers see but a limited variety of swamps and types of soil. Their opinions upon drainage are. based on traditional ideas of size of drains, 2 feet x 1 foot or 4 feet x 2 feet or even larger being conside.red necessary from an agricultural point of view to dry up an area effectively. These opinions arise. from lack of experience in juaging between swampiness due to a high water table. and that due to collected surface water lying on a heavy clay soil. On undulating land in the foot-hill districts there is 1J8\,\8 lly li ttle difficult.y in finding a sufficient fall to construct a reftsona bly well gradl3d drain sufficiently narrow and deep to carry off the normal flow of water at a speed th at will effectiveh­prevent the breeding of Anopheles larval . It is therefore desir­able that the estate medi(',al practitioner should have his own anti-malarial gang : a lines .. dresser" and two coolies are usually sufficient. They can be put on to such work as t he. medical practitioner con8i.ders necessary and can be inst,ructed to carry out the work in the manner he thinks best. Conflict of views can thereby be. largely avoided and increased flexibility of action gained. '1'hi8 freedom of action is most important when an outbreak of malaria threatens.

RAVINES .

The word "ravine" is widely used in this country to mean any form of water course in undulating or hill land. Any valley, small or: large, steep or shallow, is thus loosely referred to as a

,7

"ravine". Any fl at water-logged area , such as i.s often found towards th e foot of the undulations or hills , form ing the outlet of a st ream, or marking th e junction of two stream s , is referred to as " swamp " . Even in hilly places where th e. valleys have be­come blocked \vi th silt, or, as frequently h appens , where the road or railway culvert s are placed too high to allow of proper drainage , the result ing collections of water are called swamps.

The so-called "ravines" vary greatly in steepness and width . The ground forming the sides and base of the valleys also shows m any types of soil : th e soil most favourable. to good .1rainage work is one contain.ing a h igh proportion of clay. The average laterite so il of this country is of t,his n ature . Several types of earth, however, m ay be encounte.red even in one ravine. P at ches of white kaolin clay, gravel or ou tcrops of rock are fre­quently m et with, while, in the fl a t areas, sand is t he cause of m any difIicult ie.s . in drain construction .

RAVrNE DRAINAGE .

The purpose of the drainage of ravines is to gi.ve ready exit tv surface water, and to control flowing water and seepages.

Assum ing a r avine of moderat e proportions in width and gradient and fre.e from veget ation, drainage call s for no parti­cular skill so long as the three principles rnen ti.oned previously are kept in min~1. In organising such work the chief difficulty in dealing with the main stream is to decide how wi.de and de.ep to m ake the drain . On inspect ion it will probably be found the stream takes many bends, either because of undulations in the land forming th e. base of the valley, or because t rees have fallen across and dammed the ravine. Conseq uently th e flow of water wi ll vary frequently, now fast through narrow passages, now slow through pools or fl at swampy pat ches. '1'here is little wonder if the inexperienced m aker of drains, after seeing much water h eld up in pools and fl at areas, concludes t ha t a wide drain will be necessa ry for satisfac tory drainage.

However, as I said before , don 't be deceived by surface water. It is the flowing water with which the drain is primari­ly being constructec1 to deal. r1'h e.re is a simple way of deciding the size of drain necessary t o curry off the normal flow of t he stream. It is as follows : examine one of the narrow parts where the stream is fl owing at a reasonable speed . Measure it" width and depth. Say , for example, it measures 2 feet wid e and 3 inches deep. The simple sum of 24 x 3 inches equals 72 square inches and represents the section of th e fl owing water . Obviously a Jll'a,il1 m easuring 9 inches wide by 8 inches deep would carry off the sam e amount of w!1ter and therefore a drain !) x 18 inches would have a free margin of 10 inches depth to spare for light rain. A drain of this type ·conforms to the third principle in being narrow and deep. I would h ere 'give a fre­quently requi.re.d warning. Don't worry about storm water. Th ere is plen ty of room above the drain level for any degree of storm water . It is only on fl at lan.d that heavy rain t ends to li e about long enough for breeding. The drainage of t rue swamps is a specia.l subject calling for collaboration with th e. engineer speciali st. FigUl'e 1 shows a jungle swamp after fell­ing and burning; Figure. 2 the sam e svvamp drain ed by a 9 x 24 inch dl'ai.J1 .

,,~:~ NEGAJlAIAALAYSIA

8

DltAJ:-rAGE OF BLUKAR-COVERED RAVIKES.

A blukar-covered ravine is one which, after the felling of t he original jungle , has been allowed to revert to secondary growth of any type of vegetation that can find a footing. A high percentage of the ,ravines in hilly land districts are. in this conJi­t ion. So long as the cover is of good growth A. rnaculatus does not breed. Th ey are , however , most dangerous potential sources of unexpected outbreaks of m(1Iaria . Trees fallin g .across them, cattle wandering into them and natives making footpaths through them may dam age the shade an d give rise to breeding places.

:B or the most part such overgrown ravines are artificial .. wamps du e to collections of sprin g and seepage water dammed by spongy masses of low growing vegetation and the trunks of fallen jungle trees . The wat l' has no chance of a proper flow through tbe mass, and the base of the ravine, being usually of heavy clay , prevents percolation to lower levels. .

'rIle treatment of such a swamp is comparatively easy provided it is t ackled with pat.ience and good sense. The water, being mainly co ll ected surface water , even though it may be ;1

couple of feet deep in p laces, w i~l run off if given a reasonable chance to do so.

'l'he first step is to have the line of the proposed dra in mark­ed out. A li tt le experience is n ecessary to do this sort of work in a ravine. covereJ with t rees, grasses and creepers th rough wh ich one cannot see more than a few yard s ahead. In the n fl lTOWer ravines it is usually poss ibl e to cut the drain straight through the centre of the ravine. But if the. ravine is, say , two or three chains or more wide it will be necessarv to cut "rentices " Hcro s it at intervals of n few chains in order to' get an idea of the ti e of the land , the nature of ob tructions and the amount of flowing water.

Havi.n.g dec ided the line along which the drain can be most ea .. ily constructed, und having calculated t he size of the drain required by the method previously desc,ribed, cut a narrow "l'ent ice", or pfl th, through the hZ'/,{ kar (secondary growth) and remove all t rees , bushes and obstructions along t he. lin e of the proposed drain , In cuttin g paths for such purposes inside swampy rav ines , it is nec ssary to keep them as narrow as possi­ble 0 that as little sun ligbt us possible can enter , . otherwise " rnacu IHtus" will soon establi sh breeding places. It might well be asked why not fell all the hlllha l' first so th at t he line of the dra in can be easily seen : this is one of the most frequ ent mistakes made by t he agriculturist . Such fe lleel blulw r has to lie for several weeks to dry before it can be cut and pile,el ~'eac1y for burn in g. Durin g this t im e it is im ­possible to get in and efficiently oil the area. Thus" macul atus" has un excellent opportunity to establish breeding places and usually makes th e most of it . Only by draining a bZttlwl'-covered swamp before felling can A . 1nnclllatus breeding be controlled.

It is my pract ice to p ut in two coolies armed only with 1iamngs-heavy Im ives-after cutting the necessary "ren t,ice" anJ marl<ing out the lin c of the drain , They are instructed to follow th e line of th e cleared "rentice" and to cut through t he grass and roots which fO~'m the base of such swamps and lift t he mass out with t heir hancls . '1'he side of the channel thus formed are soft and coll aps ible. It may tb erefore be necessary to keep R labourer going over t he line every few days, pressing the