CO885-(25-26) — Page 117

213

115

Table II. (b)-Catch in mainland districts where female percentage

averaged 45 or over.

(b) Relative inactivity of females as the explanation of low female percentage in catch of flies.

Table III. Showing reduction in density of fly on Tavu Island due to catching experiment, and indicating presence of inactive females. Table IV-Showing reduction in density of fly on Lula islet, as result of

catching experiment, and proving the presence of inactive females. (c) The percentage of females among caught flies as an index to food-supply. Table V-Showing increased activity of females on Lula islet, which may

have been due to presence of fly boys frightening away host animals. Details of experiment proving that increased activity of females and higher

percentage of females in catch may be caused by banishing wild hosts. (d) Conclusions. Speculations concerning the real ratio between the sexes, which is impracticably measurable.

II.-The variable behaviour of Glossina palpalis toward Man and Domestio

Animals, as correlated with variations in Sex Ratio:-

(a) Methods used for estimating the density and the economic status of Glossina

palpalis:

(1) Estimating variations in density of fly;

(2) Estimating variations in economic status of fly as determined by food-

supply.

Table VI. Showing extremes of variation in economic status of Glossina palpalis as determined by food-supply and indicated by female per- centage.

(b) Correlation between high percentage of females artificially produced and persistence of attack upon man.

Table VI. (a). Showing great increase in persistency of attack upon goats and pigs caused by driving away reptilian hosts, and correlated with increased percentage of females.

Notes on changed behaviour of flies toward man under identical conditions. Note on changed behaviour of flies following accidental banishment of wild

hosts.

(c) Correlation between high percentage of females naturally prevailing and persistency of attack upon man.

Table VII.-Illustrating above.

(d) Experiment to test variability in behaviour of fly toward man on Konu and Damba Islands, as correlated with variations in abundance of host animals and in female percentage.

Table VIII.-Illustrating above.

shore.

(e) Conclusion.-The female percentage may be a very valuable index to the chances favouring transmission of human disease by Glossina palpalis. III.-The Longshore Movements of Glossina palpalis:- (a) Experiment to test the movement of flies along

Crocodile Point,' Bulago Island, showing Table IX-Catch of fly at

existence of rapid movement of flies. (b)_ Movements of the sexes along shore.

Table X.-Catch of fly for consecutive hours at point along shore, showing reduction in local density of males due to catching, but not of females, and demonstrating more rapid movement of females than of males. Table XI. Catch of fly for consecutive hours at a point on shore of island where food was deficient, showing equal activity on the part of both sexes and equal rapidity of movements along shore.

(c) Effect of shelter (or type of vegetation) upon longshore movements of fly

and upon percentage of females in the catch.

Note and data, showing extreme variation in percentage of females in catches made at the same locality but inside and outside the reed fringe along shore, respectively.

Note and data. showing extreme variation in percentage of females in catches made at the same locality, but at well sheltered and poorly sheltered points, respectively.

Table XII. Catches of fly at various points on island of Kimmi, showing effect of variable character of shelter on movement of sexes and female percentage.

Graph I. Same data graphically presented.

Graph II.-Fly survey of 3,000 yards of shore on Bugaba Island, showing effect of shelter on movement of sexes, density of males, and percentage of females.

Graph III.-Fly survey of five miles of shore in Lutoboka Bay, Bugalla Island, showing effect of shelter on movements. density, and ratio of

sexes.

(a) Notes on survey of Lutoboka Bay, as shown graphically in Graph III. (e) The maximum percentage of females as brought about by segregation through more active movement.

Table XIII-Catch of fly along three-and-one-third miles of shore near the mouth of Mujuzi Creek, in Buddu, showing extremely high female per- centage, due in part to scarcity of food and in part to segregation of the sex througa more extended movements in search of food.

() Summary and Conclusions.-The full significance of the phenomenon of sex disparity: the average female percentage for a region or district is an index to quantity of available food, therefore to the persistency with which flies will attack man and domestic animals, and therefore to the chances favouring trans mission of disease. The female percentage, taken point by point within a given region or district, is indicative of character of shelter and general attractiveness of environment at the points in question. It is correlated with concentration of males at points where environment is most attractive, and with movement of hungry flies (females predominating) into unattractive environments in search of food. IV.-The Assembling Habit of Glossina palpalis. Movement of flies stimulated by

the masculine instincts :-

(a) Observations.-The "following swarms" of flies consist largely of males, which are attracted to host animals not out of desire to feed, but in order to meet the females. Notes on following swarms accompanying various hosts as follows:-

"

On Varanus: Kitobo Island, 3rd December, 1913. On Varanus: Nsadzi Island, 3rd February, 1914. On Situtunga: Bugalla Island, 23rd November, 1914. On Situtunga: Damba Island, 13th September, 1915. On Varanus: Manene Island, 16th March, 1915. On Situtunga: Bugalla Island, December, 1914.

"}

(b) Conditions under which a following swarm will collect only on favoured hosts, which are complacent under attack.

V-The effect of Open Water and Humid Conditions on Range of Fly:-

(a) Observations showing normal range inland from water: only a few hundred yards.

(b) Effect of marshes and rivulets on range.

Table XIV-Showing complete lack of correlation between inland range

of fly on islands and presence of marshes or rivulets.

(c) Effect of food-supply on inland range.

Table XV-Showing abnormal extension of range inland correlated with

presence of game in unusual quantity.

Table XVI.-Showing complete correlation between inland range of fly on

islands with abundance of host animals.

(d) Inland range as affected by Varanus. Notes and observations on various islands.

Notes and observations at various

(e) Extension of range inland on mainland.

points on extension of range due to bush pig or bush buck.

(f) Dispersion of fly inland from lake shore along small rivers.

Graph TV-Showing dispersion inland along banks of Mujuzi Creek as compared with dispersion along lake shore on either side of mouth of creek.

(g) The sex ratio at points where range is extended inland. (Nearly always the female percentage is lower than on the shore, indicating more attractive environ- ruent and better food-supply.)

P 2

PUBLIC RECORD OFFICE

Reference :-

TILLEC885/25

PUBLIC RECORD OFFICE, LONDON

ALLY WITHOUT PERMISSION OF THE BE REPRODUCED PHOTOGRAPHIC-

COPYRIGHT PHOTOGRAPH-NOT TO