N. H. SWELLENGREBEL AND J. M. H. SWELLENGREBEL-DE GRAAF 1 93 
may be found breeding in all imaginable collections of water (salt, brackish, 
fresh, dirty, clean, running or stagnant). In the absence of natural breeding 
sites, it will thrive in artificial ones (water in boats drawn up on shore, water 
in coconut shells, etc.) as readily as will Culex fatigans or Stegomyia scutel- 
lans. The larva of M. indefinita behaves in a similar way, except that it is 
not often found in distinctly brackish water: we even found it breeding in 
saucers, of water, placed to prevent the access of ants. C. kochii and^ N. 
punctulata (tesselata ), though somewhat more fastidious, distinctly belong to 
this group. 
Still there are limits to the breeding possibilities of these species. In water, 
highly polluted with faecal matter, the larvae are absent; we once found 
.. . — . ^ ^ ^ o species were found to be 
very limited in the choice of a breeding-place. The first of these, S. aitkenii, 
occurred in swiftly running mountain streams in 97 per cent, of cases. In 
this restriction of habitat it compares with what is stated concerning N. 
maculatus. If S. aitkenii proves to be a vector of malaria, this knowledge 
may be put to practical use. The second species, M. ludlowi, has at present 
a more practical importance. It is found in two morphologically indistinguish¬ 
able varieties, which, on physiological grounds, must be dealt with separately. 
Ihe first or httoral form, is almost confined to salt or brackish water along 
the coast, especially in places where algae (Enteromorpha, Cladophora, 
Lyanophyceae) and other filamentous aquatic plants, like Najas, are growing. 
In such breeding-places, during the dry season, June-September, the larva 
may be found in immense numbers. It is occasionally found in fresh water, 
but even then it is limited to the coastal zone, rarely penetrating inland for 
more than 4 km., except in such places where brackish water is to be found 
far inland (tidal rivers). Where the tide runs in and out freely, as in mangrove 
swamps it appears to be absent; but if human intervention,' by the construc- 
lon of bunds, locks, etc., creates obstacles to the tidal water, the conditions 
change; grass begins to grow between the roots of the rhizophores, and larvae 
of M. ludlowi soon begin to appear. A place comparatively healthy before 
he commencement of the operation is thus rendered highly malarious (see 
iable V). 
The larvae of the inland form of M. ludlowi are to be found only in the 
narrow valleys among the mountains running parallel to the west coast of 
feumatra, where they occur in numerous fish-ponds. These ponds are almost 
he only breeding-places of this species, but it may breed in rice-fields also, 
though rarely. In one of the valleys, an extensive lake (Manindjau) is situated, 
a ong the banks of which this form is common among the rushes and algae’ 
a Irequent habitat of many Anopheline larvae (Table V). 
13—2 
