534 
attacks. It is difficult to discover how far these forms were common 
in Rome, owing to the way in which Roman writers on medicine 
repeated the remarks of their Greek predecessors, but we are dis¬ 
tinctly told that the semi-tertian was extremely prevalent. 
Both in Greece and in Italy the geographical conditions favour 
the development of malaria, as is proved by the prevalence of the 
disease in modern times. But, for the present discussion, attention 
must be paid to the neighbourhood of Athens and Rome. Near 
Athens were two streams, the Cephisus and the Ilissus, which, by 
partially drying up in summer, favoured the rapid growth of the 
mosquito. Near the Piraeus was a marshy district which must have 
proved a continual focus of malaria. Besides these natural breeding- 
places, the cisterns in which the Athenians kept their water seem to 
h.i\e harboured mosquito larvae. Aristotle tells us that they often 
contained the larvae of an insect which was, in all probability, 
C hironomus. The streets also of the city were very muddy in wet 
weather, so that puddles in out-of-the-way places were certainly 
numerous. I may refer the reader to four able articles on malaria 
in Greece, by Drs. Savas and Cardamatis, which appear in the viiith 
v 190;) volume of Atti della Societa per gh Studi della Malaria. 
geographical conditions are shown to be well adapted to the 
growth of the mosquito. 
T h^fl 1 * 01 , 116 Simikr COnditions Prevailed. The inundations of the 
00 ed a considerable portion of the land near its banks, which 
f , notoriousl y unhealthy. 1 he streets seem to have been as 
mu y in wet weather as were those of Athens, while the atrium of 
• ; , oma n house contained a pool of rain-water which collected 
to the impluvium through a hole in the roof, intended no doubt, at 
l ° ICt ° Ut the smoke froni the household hearth. 
. , . * C c notlce d that just at the time when malaria appears 
sLerelirTTu endemiCin Att ' Ca and Ital N ( 4 *> and 200 B.C.), 
proper control f * C ° Untry Waste> and Prevented for many years the 
Ihe'spr^ o^^ atln " StrCamS and CaaaIs ‘ ^is would favour 
