250 ABSTKACTS: TECHNOLOGY 



TECHNOLOGY. — The 'properties of Portland cement having a high 

 magnesia content. P. H. Bates. Bureau of Standards Technologic 

 Paper No. 102. Pp. 42. 'January 19, 1918. 



The question of the maximum amount of magnesia allowable in 

 Portland cement is one of the most interesting encountered in the study 

 of this complex material. There is still much diversity of opinion. 

 Magnesia in amounts not greater than 8 per cent is believed by many to 

 be harmless, whereas others consider amounts greater than 4 per cent 

 injurious. 



It was considered very desirable to investigate the subject because 

 failures of mortars and concrete were attributed to high magnesia 

 content and especially so since all investigations to date have been 

 somewhat at fault. 



A series of cements have been burned in the experimental rotary 

 kiln of the Bureau in which the limestone used in the raw material 

 was replaced in part or in whole by dolomite. Cements were thus 

 obtained in which the magnesia content varied from 1.77 per cent to 

 25.53 per cent. The results obtained show that cements when contain- 

 ing not more than approximately 8 per cent magnesia will produce con- 

 cretes with satisfactory strength at the end of one and one-half years. 

 At this amount of magnesia, monticellite and spinel (constitutents not 

 present in cement of lower magnesia content) appear; and those cements 

 seem to hydrate with a large increase in volume. 



P. H. B. 



TECHNOLOGY. — The determination of absolute viscosity by short- 

 tube viscosimeters. Winslow H. Herschel. Bureau of Standards 

 Technologic Paper No. 100. Pp. 55. November 9, 1917. 

 The Engler and the Saybolt Universal viscosimeters, which are the 

 instruments usually employed in the oil trade, have such short outlet 

 tubes that the equation for the flow through long capillary tubes is not 

 applicable without correction factors. The literature has been care- 

 fully reviewed and further experimental work has been done. The 

 conclusion is reached that water is not a suitable liquid for use in finding 

 the relation between viscosity and time of discharge for short-tube 

 viscosimeters, and that Ubbelohde's equation, and all others based 

 upon it, are seriously in error. 



W. H. H. 



