April 8, 1920] 



NATURE 



85 



As regards the measurement of diameters, although 

 an ordinary micrometer will suffice for the external 

 limension, it is not suitable for determining either 

 he effective or the core diameter. Before testing the 

 . iTective diameters it is necessary to know the errors 

 of the pitch, in order that the appropriate reductions 

 may be made in the maximum limit of tolerance. A 

 micrometer with point contacts should never be used 

 alone for the effective diameter, as it bears only on 

 projecting parts, and, further, the points wear down 

 quicklv. Even when new, its contacts for screw- 

 threads rarely have the correct angle. It is, however, 

 a useful check on results obtained by the aid of wire 

 contacts, especiallv for investigating anomalies which 

 may be apparent in these results. 



Mr. Powell has made a special study of wire con- 

 tacts for testing effective diameters. ' He employs 

 systematically two series of wires for each pitch of 

 -rrew. One series is such that the wire bears exactly 

 in the^ theoretical effective diameter of a perfect 

 screw ; in the other series the wire bears on the sides 

 of the screw not far from the outer edge, but so as 

 to avoid the rounded-off part in Whitworth threads. 

 The correct diameters for the series are calculated 

 from a simple formula. The wires, which are finished 

 by grinding, must be perfectly cvlindrical, and their 

 diameters have to be ascertained to an accuracy of 

 oooooi in.; any error in the diameter of the wire 

 is multiplied threefold in the result obtained for the 

 effective diameter. Thev are made of hardened steel, 

 and are about 2 in. in length, or longer for verv large 

 screws ; but their exact adjustment is only necessarv 

 over about half an inch in the centre of the length. 



The verification of internal screws may be effected 

 simply by emnloving either an external screw havinp- 

 the correct dimensions for external, effective, and 

 core diameters, or a screw correct for maximum and 

 effective diameters, but slijjhtlv small for core dia- 

 meter. A plug is also used in this case for verifying 

 the minimum diameter of the internal screw. If 

 these gauges enter the nut, the test is re.<*arded as 

 satisfactory ; but, in reality, this is not aUvavs the 

 case, as the external screw may appear to jjive a good 

 result even if it bears on only one diameter of the 

 nut ; while the other diameters may be far outside the 

 limit of tolerance. A large number of different gauges 

 are necessary in order to verify separately every 

 diameter of an internal screw. 



Mr. Powell has drawn up a list of equipment 

 required in the verification of screw gauges. This 

 includes a pitch-measuring machine, an apparatus for 

 obtaining casts of internal screws, a projecting ar- 

 rangement, an instrument for measuring the three 

 characteristic diameters of external screws, a collec- 

 tion of suitable wire contacts, small triangular prisms 

 for verifying core diameters, standardised micro- 

 meters. Johansson gauges, and a complete set of 

 standards for measuring diameters, pitch, and form 

 of internal threads. 



.A theoretical explanation of the principles of the 

 methods employed would have been of interest. In 

 its absence some doubts arise, for example, as to the 

 practical value of profiles of screw-threads projected 

 on screens. Attain, the contacts of small cvlindrical 

 wires on the helicoidal surfaces of threads cannot be 

 rpfjarded as the same as that of a circle and two 

 intersecting straight lines, although the formulae 

 employed, which are stated without proof, apnear to 

 be founded on a consideration of this kind. In con- 

 clusion, Mr. Powell's methods are bv no means 

 entirely novel, btit thev were verv successful during 

 the war. and will no doubt be found instructive by 

 all those en£?a£*ed in the manufacture and verification 

 of screw gauges. 



NO. 2632, VOL. IO5I 



The Composition of Salvarsan. 



AX/" HEN salvarsan was first introduced for use in 

 * '^ medicine the German manufacturers statecJ that 

 it contained "about 34 per cent, of arsenic," which is 

 the percentage calculated for a pure dihydroxydiamino- 

 arsenobenzene dihydrochloride, C,jH,20,N2ASj,2HCl. 

 This statement was afterwards altered to "the arsenic 

 content of the preparation corresponds to the formula 

 C,,Hj,02N,As,,2HCl,2H30 as a result of Gaebel's 

 observation that the drug loses 76 per cent, by weight 

 on drying, and contains only 31-5 per cent, of arsenic." 

 Last year Kober, in the United States, ventured the 

 opinion that the combined solvent in salvarsan is not 

 water, but methyl alcohol, and suggested that the 

 latter might be the cause of variable toxicity in sal- 

 varsan^ — a suggestion which is rather far-fetched in 

 view of the fact that, even on Kober's assumption, a 

 maximum normal dose of salvarsan could contain only 

 004 gram of methyl alcohol. 



This and other questions connected with the com- 

 position of salvarsan have been investigated in the 

 Wellcome Chemical Research Laboratories, and in a 

 paper contributed to the meeting of the Chemical 

 Society on March 18 Messrs. Fargher and Pyman 

 showed that the combined solvent in salvarsan is 

 water; and though small quantities of methyl akohol 

 may also be present, due to the use of this alcohol in the 

 liquid from which the drug is precipitated, the arnount 

 never exceeds 14 per cent., and is frequently nil. It 

 was also found that the sulphur always present in 

 commercial salvarsan as a result of the use of sodium 

 hyposulphite as a reducing agent in its preparation, 

 occurs in at least two forms: (i) as a sulphaminic 

 acid, probably "salvarsan" monosulphaminic acid 

 hydrochloride,' and (2) attached directly to arsenic; 

 whilst a third portion may be in physical association 

 with salvarsan, which has certain colloidal properties. 



These results support the conclusion expressed in the 

 recent Special Report (No. 44) of the Medical Research 

 Committee, that though salvarsan is not a chemically 

 pure substance, there is no known chemical imputity 

 with the presence or proportion of which its varymg 

 toxicity can be brought into relation. In this con- 

 nection it is interesting to note that a specially pure 

 salvarsan free from sulphur, prepared by Messrs. 

 Fargher and Pvman, was tested by the Medical 

 Research Committee and shown to be more than 

 normally toxic. Chemical testing alone is, therefore, 

 insufficient to determine whether any particular batch 

 of salvarsan is suitable for medical use, and it i.s on 

 this account that the Medical Research Committee 

 has elaborated the system of biological testing, 

 described in the Soecial Report already referred to, 

 to control the issue of salvarsan in this country. It 

 is satisfactory that the Committee is able to report 

 that from the point of view of permanence of effect, 

 the British and French salvarsan preparations are 

 therapeutically as good as the German. 



University and Educational Intelligence. 



'The governors of the Huddersfield Technical Col- 

 lege have received a gift of 2oooi. from Mrs. Mary 

 Blamires, widow of Alderman Joseph Blamires, m 

 memory of her late husband, himself a former 

 student, and afterwards a governor, of the college. 

 The scholarship is to be used for the promotion of 

 research in chemistry. 



The headquarters of the Yorkshire Summer School 

 of Geography, now being organised bv the University 

 of Leeds, will this vear be the County School, Whitby, 



