ALCOHOL FDELS. 145 



•engines. It consists of a one and a half kilowatt single cylinder 

 -4 cycle internal combustion engine with generator combined, work- 

 ing through an air resistance so that any desired load can be 

 obtained. It was my intention to have the work done shown on 

 a recording wattmeter, but this has not been delivered yet, so I 

 have had to be content with an ordinary voltmeter and recording 

 ammeter. Although this is rather a crude testing set I have got 

 some very interesting data, which, if not strictly accurate, are 

 much more accurate than any tests which could be carried out on 

 Ihe road. 



According to the calorific value Natalite has only 72% the 

 value of petrol. The value is approximately 12,000 B.T.TJ. I 

 have not been able to have it taken in a calorimeter, but have taken 

 petrol as 18,000 B.T.U. and alcohol as 11,500. 



I have had to calculate the calorific value of ether from 

 Dulong's formula corrected by Redtenbacher, which gave an 

 approximate value of 14,000 B.T.U. for ether, corresponding to a 

 calorific value of 12,000 for Natalite. 



In the running tests with the testing set, an average of five 

 different runs was taken with the following results: — 



At full load Natalite gave 75-6 % compared with Shell petrol. 



At 2/3 load Natalite gave 79-65% compared with Shell petrol. 



At 1/2 load Natalite gave 80-27% compared with Shell petrol. 



At 1/3 load Natalite gave 86-85% compared with Shell petrol. 



This gives an average on all loads of 80-59%. 



The higher efficiency at 1/3 load is rather extraordinary. I 

 think that the effieiencies at all loads would be increased by sup- 

 plying hot air to the carburetter so as to vaporize the fuel better. 

 This is the more necessary as the engine has a fairly long induc- 

 tion pipe. However, the conditions were similar to those met 

 with in the majority of motor car engines. 



It is rather difficult to explain the difference of 8% efficiency 

 over the calorific value of the fuels. The thermal efficiency of the 

 engine working on the alcohol fuel is certainly higher than that 

 of the petrol engine on account of the lesser quantity of heat 

 lost in the exhaust owing to less air being required for combustion, 

 but this does not account for all the difference. In each case the 

 needle of the carburetter was closed until the engine started to 

 miss fire and then opened again very gradually until the missing 

 stopped. The same carburetter setting was kept for all loads. 

 These results agree very nearly with those that I get on my car 

 on the road, viz., a mileage of practically 80% of that of petrol. 

 The efficiency at the lower loads is very good when it is considered 

 how much running about is done at one-third or half load. In 

 my opinion an engine as built for petrol is very suitable for alcohol 

 fuel. One often hears people extolling a new alcohol fuel, saying 

 that the compression of the present engines should be increased 

 and wonderful results would follow. I have tried increasing the 

 compression to llOlbs., but only got slightly better results. I 

 think that the compression would have to be raised to about 

 1501bs. per square inch or higher to get really good results. The 



