21 



conditions permitting the malarial organism to develop in the mos- 

 quito, but the author considers that many of these primary cases have 

 remained latent after infection in the preceding autumn. In autumn 

 Anophelines are very abundant indoors, and in September and early 

 October the temperature permits the cysts to mature in the mosquito. 

 In spite of this there are few primary cases in October and November. 

 The same reason that prevents relapses in late autumn and winter 

 postpones first attacks until the spring. 



Assuming an autumn infection in many spring primary cases, many 

 of the latter may be expected to originate in families in which malaria 

 occurred after 1st September in the previous year. This seemed to hold 

 good in 20 out of 146 primary cases noted before 1st June. On 

 dividing the families concerned into two groups, in which malaria had 

 occurred before or after 1st November, the latter group seemed con- 

 nected with 2 only of the 20. It is therefore probable that mosquitos 

 are not easily infected after 1st November. It is possible to find a 

 like relation between the spring primary cases and early spring relapses 

 in the same families, and this relation is a feasible one as regards time. 

 Investigations indicated, however, tnat persons harbouring latent 

 malaria in spring do not give Anophelines an opportunity to become 

 infected. 



In the second part of the paper Dr. Swellengrebel points out that, in 

 order to state when man becomes infected, it is important to ascertain 

 the occurrence of infection in Anophelines and compare this in order of 

 date with the human infection. 



For this purpose Anophelines were caught in houses where mialaria 

 had occurred. The results, which are admittedly incomplete, show that 

 the maximum infection of these mosquitos occurred in late autumn. 

 At Amsterdam no infection was found before August, and the maximum 

 reached was 6-6 per cent, in November. At Nieuwendam the maxi- 

 mum was in October and November, but was below 2 per cent., while 

 infections occurred in March and June to the extent of 1 per cent, 

 in each month. This difference between Amsterdam and Nieuwendam 

 may be explained as follows. In summer the mosquitos leave buildings 

 to oviposit and are unlikely to return to the same ones. Stables and 

 cowsheds are very attractive, and infected mosquitos are difficult to 

 find among the large numbers that occur in them ; their numbers 

 are also constantly being added to by newly emerged individuals. 

 The chance of finding infected mosquitos was therefore very small 

 in summer, but greater at Nieuwendam than at Amsterdam as there was 

 more malaria in the former locality. In winter Anophelines do not 

 oviposit, and therefore come out of buildings much less and travel 

 shorter distances. They thus remain congregated indoors and infect 

 the neighbouring dwellings. At Nieuwendam, however, stables and 

 cowsheds are interspersed with dwellings, and there is a chance of their 

 sheltering in the former after even a short excursion out of doors, 

 so that they do not congregate inside malaria-infected houses to the 

 same extent. 



As regards the development of winter infections. Dr. Swellengrebel 

 states that in December, January and February he found four mos- 

 quitos with sporozoites in the salivary' glands ; the infection can 

 therefore develop in winter. Furthermore, in 7 out of 18 infected 

 mosquitos caught in November the state of development of the oocysts 

 pointed to the maturing of sporozoites. Studies of development 

 showed that after October no smaller half-grown oocysts were found, 



