otjr climate — the liot dry air of our long South Australian 

 summer. 



Of the total number of deaths of infants in the province, the 

 following proportions occur in the several months: — Most in 

 March, 124; December, 121; January, 11*3; February, 10-8; 

 April, 10"3 ; May, 8'3 ; November, 7 ; August, 5'8 ; June, 5*7 > 

 October, 5'5 ; September, 5"5 ; July, 5'3. 



Five summer months head the list and five winter ones close 

 it. As many babies die in four summer months — viz., January, 

 February, March, and December— as in seven others. Twice 

 as many die in January, February, March, and December as in 

 June, July, September, and October. 



Comparing the mortality with the temperature,- we find the 

 highest average temperature is in January. In this month the 

 mortality is extremely high. In February both the tempera- 

 ture and mortality are slightly less. In March, the temperature 

 is a little lower, whilst the rate of infant mortality attains its 

 highest point. This must be either because the long-continued 

 heat has worn . the children out so that they cannot withstand 

 the March temperature, or because the houses have become 

 so thoroughly heated that though the air is not quite so hot, 

 the temperature inside the houses and to which the infants 

 are exposed, is just as high and even more close and oppressive. 



From March, the temperature falls very rapidly during April, 

 May, and June, and reaches its lowest point in July, and as 

 it falls the rate of infant mortality follows pari passu in a 

 remarkable way. It also reaches its lowest point in July. 



In August, the temperature rises a little, and with it rises 

 the rate of mortality. In September and October the tempera- 

 ture still slowly rises, but the rate of mortality remains almost 

 stationary and not at all excessive. 



In November and December, the temperature increases very 

 rapidly, and with it the deathrate increases by " leaps and 

 bounds," and in December becomes very high indeed. 



From this sketch we conclude that the rate of mortality 

 depends to a very large extent upon the height of the ther- 

 mometer, and this with a regularity that is quite surprising, 

 and to my mind proves that (excluding epidemics) the height 

 of the thermometer during any given year will be the guide as 

 to the amount of danger to which infants are exposed, provided 

 that we make allowance for the humidity of the atmosphere. It 

 is also noticeable that the thermometer begins to fall a little 

 before the deathrate ; it passes, but leaves a dreadful train to 

 follow, and it begins to rise again a little before the deathrate 

 increases, but as it rises the deathrate soon follows, showing 

 that it is not exactly heat that kills, but continued heat. 



With eespect to Humidity. — Upon this diagram I have 



