422 FAMILY: TRYPANOSOMID^ 



repeated examination of the faeces of fleas to be made, so that natural intestinal 

 infections can be excluded before employing them for feeding experiments. On 

 account of these and other incomprehensible statements, it seems impossible to 

 accept Basile's claim that he has demonstrated the flea transmission of Mediter- 

 ranean kala azar. Pereira da Silva (1913, 1915) conducted a very careful series of 

 experiments, employing Noller's method amongst others, with a view to deter- 

 mining the possibility of L. donovani developing in the human and dog flea. He 

 could obtain no evidence whatever of such a development, and came to the con- 

 clusion that the flea is not the transmitting agent of kala azar. Basile (1914, 

 1914rt) attempted to explain the negative results obtained by other workers by 

 assuming that certain meteorological conditions existed during his experiments 

 which were absent during those of other observers. It is of interest to note here 

 that the writer (1912c), experimenting with fleas by Noller's method, could obtain 

 no evidence of the development of L. tropica in these insects (see below). Xicolle 

 and Anderson (192.3, 1924) have published an account of most careful attempts to 

 transmit kala azar to ten dogs by means of numerous dog fleas which had fed upon 

 known infected animals. The exposure to the fleas lasted from three weeks to seven 

 months, but in no case did an infection result. Two of the dogs were made to 

 swallow 510 and 410 fleas. Every source of fallacy was excluded, and it is rightly 

 concluded that the experiments lend no support to the flea transmission hypothesis. 



Mosquitoes. — Franchini (1911a, 1912) allowed Anopheles macuUpennis to feed on 

 cultures of L. donovani. The parasites persisted in the gut up to twenty-four hours, 

 and leishmania forms were passed in the faeces. The same mosquitoes were fed on 

 spleen puncture material from cases of kala azar. Tieishmania were ingested, and 

 persisted up to forty-eight hours. After thirty hours large round forms were present, 

 while at the end of forty-eight hours a flagellate leptomonas form Avas found. He 

 claims to have controlled his experiments by numerous dissections of mosquitoes 

 not fed on leishmania material. Flagellate infections of A. macuUpennis are, how- 

 ever, very common. No evidence of the possibility of transmission was produced. 

 Patton (1907 and 1912fl.), in India, could obtain no evidence of any development 

 of leishmania in Culex pipiens, A. Stephens i, and Stegomyia sugens, but natural 

 flagellates were found in some of these. Similarly, Mackie (1915) was completely 

 unsuccessful with 266 culex and eighteen anopheles. 



Lice. — Patton (1907-1912rt) obtained negative results after feeding lice {Pedicnlus 

 capitis and P. vestimenti) on kala azar cases in the blood of which leishmania 

 occurred. Mackie (1915) fed large numbers of lice on kala azar patients, and, 

 furthermore, dissected larger numbers collected from cases without finding any 

 trace of leishmania in them. In all, over 3,000 lice were thus examined. 



House Flies. — Patton noted that L. donovani degenerated very rapidly in the 

 intestine of the house fly. 



Ticks. — Patton, working with Ornithodorus savignyi, could obtain no evidence 

 of development of L. donovani. Basile (1910a) and Marshall (1912) likewise had 

 negative results with ticks. 



Sand Flies. — The recent experiments of Knowles, Napier, and Smith (1925) have 

 directed attention to sand flies of the genus PhJebotomus. They point out that Sin ton 

 has informed them that in India the distribution of Phlebotomus argentipes coincides 

 with that of kala azar. An investigation of this fly in Calcutta has shown that 

 twenty-five out of fifty-six female flies bred in the laboratory contracted a leptomonas 

 infection after feeding on kala azar cases. Bred flies, forty-six in number, fed on 

 control cases acquired no such infection, while 407 wild flies (317 ? and 90 (^) also 

 showed no infection. Similarly, 210 wild P. minutas were uninfected. Experimenting 

 with P. minutus, it was found that this flv would not feed on man. One hundred 



