Fly No. 3. No parasites. 
Fly No. 4. No parasites. 
Fly No. 5. Contained a aumber of small forms which are 
characteristic of the post-flagellate stage of Hcrpetonion- 
as muscae domesticae. 
■ Fly No. 6. Contained a number of adult flagellatjs. 
Fly No. 7. Contained a number of adult flagellâtes. 
Fly No. 8. Contained a large number of adult flagellâtes. 
Fly No. 9. Contained a large number of adult flagellâtes. 
Fly No. 10. This fly had not fed and therefore did not contain 
any parasites. 
Fly No. II. No parasites. 
Fly No. 12. Contained about a hundred adult flagellâtes. 
Fly No. 13. Fly had not fed, not flagellâtes. 
Fly No. 14. No pai'asites. 
Fly No. 15. No parasites. 
Fly No. ib. Contained many adult flagellâtes. 
Fly No. 17. Contained many adult flagellâtes. 
Fly No. 18. No parasites. 
Fly No. 19. No parasites. 
Fly No. 20. Contained many adult flagellâtes. 
Fly No. 21. Contained many adult flagellâtes. 
tly No. 22. No parasites. 
Tlie last fly was clissected at 5 P. M., so that the parasites 
coLild not hâve originated from cysts ; the longest interval between 
the feed and the time of examination of fly 22 was two and a half 
hoLirs. There can therefore be no doiil^t whatever that the paras¬ 
ites seen were ingested as flagellâtes. In each case the whole 
alimentary tract of the fly was dissected ont in 0,6 % saline solu¬ 
tion and examined with a Zeiss D objective under a cover glass. 
The above experiment therefore conclusively proves that at least 
50 % of clean hatched flies may ingest the flagellated forms of 
Herpetomonas miiscac domesticae when fed for the first time 
on a piece of spleen on which infected bazaar flies had previously 
fed. I need hardiy say that the goat’s spleen contained no para¬ 
sites of any description. 
A long sériés of experiments were next carried ont w-ith clean 
bred flies w'hich had been infected once as described above and 
as thev were subsequently fed on clean goat’s spleen pulp I was 
able to follow on each succeedmg day the varions changes iinder- 
gone bv the parasites ingested at the first feed. Each experiment 
was repeated several times so that beyond slight variations in 
the number of parasites encountered precisely similar results were 
obtained. I will first describe what happens to the long flagellâtes. 
On the second day a gieat increase in the number of these para- 
