November 26, 1908] 



NA TURE 



103 



which his former students and friends expressed their 

 admiration for de Vries and his work, on the occasion 

 of the twentj'-fifth anniversary of his professorship at 

 the University of Amsterdam. On this occasion de Vries 

 <vas presented with a considerable sum of money, 

 which was expended in the erection of a vast green- 

 house, which enabled him to defy the climate 

 of Holland, against which he had been con- 

 tending for many years with anything but complete 

 success. 



The Experimental Garden at Amsterdam, as it 

 now stands, is the result of an attempt to perfect a 

 method of observing the origin of species. The 

 success of this attempt will rank as one of the greatest 

 achievements in biology. 



THE SURVEY OF AFRICA.' 



THE fifth volume of the account of the geodetic 

 survey of South Africa, executed under the super- 

 vision of Sir David Gill, has now been issued. With 

 the four volumes previously published the descrip- 

 tion of the whole work, from the southernmost point 

 of the continent up to the Zambezi River, is thus 

 completed. A sixth and final volume is promised, 

 which will comprise that portion of the thirtieth 

 meridian arc done by Dr. Rubin, carrj-ing the survey 

 northward from the Zambezi to a point 70 miles 

 south of Lake Tanganyika. This will therefore round 

 off the South African part of this great undertaking, 

 the first idea of which was originallv conceived by 

 Sir D. Gill in 1879. To him, together with his able 

 lieutenant. Colonel Sir W. G. Morris, the credit of 

 thus carrying through this immense task, in face 

 of many political and financial difficulties, must be 

 ascribed. 



The present volume is replete with interest both 

 to the scientific surveyor and to the student of public 

 policy on the questions of survey and map-making. 

 The main interest naturally centres about the intro- 

 duction by Sir D. Gill, and the introductory report 

 on the trigonometrical sur\ey of the Transvaal by 

 Sir W. G. Morris. The former gives a succinct 

 history of the triangulation of South Africa, recapitu- 

 lates the now well-known proposal to extend the 

 thirtieth meridian arc through the continent, and con- 

 cludes with a detailed resume of the negotiations 

 T^etween the Imperial Government and the colonial 

 authorities for the formation of a federal survev 

 department. These extended, with intermissions, from 

 iqoi to 1004, and finallv ended abortivelv, one colony 

 after another deciding that they could not afford the 

 expenditure necessary for the construction of an ac- 

 curate map of their territory. The expenditure ulti- 

 mately and implicitly involved by the existence of 

 inaccurate maps or by the complete non-existence of 

 any maps at all, being an item which does not come 

 on the estimates for the vear, is, we must perforce 

 conclude, a subject of little concern to the politician. 

 Otherwise, unless we are to assume that public 

 memorv is so short that a period of three or four 

 years is sufficient to drive the most striking events 

 out of mind, it is difficult to see how one of the main 

 object-lessons of the .South African war, the extreme 

 costliness of had maps, should have been so soon 

 and so completely set aside. Sir D. Gill's account 

 not unnaturally gives special prominence to those 



1 Geodetic Survey of South Africa. Vol, v. Report", on the GeoHe*ic 

 Siir\'ey of the Tmnsvaal and Orange River Colony, executed by Colonel Sir 

 W. G. Morris, K.C M.G., C.B., and of it5 connection by Cant. H. W. 

 Gordon, R.E., with the Geodetic SMr^•ev of Southern Rhodesia, with a 

 preface and introduction hy Sir David Gill, K.C.B., F.R.S. Pp. .\xx-vii + 

 463-I-16 plates ; 6 tnaps. (London : Harrison and Sons, 190S.) 



NO. 2039, VOL. 79] 



parts of these proceedings which took place in Africa, 

 with which he was directly concerned. The result 

 is that he does less than justice to the part played 

 by the War Office, and is apparently unaware that 

 the proposal to carry out a complete survey of South 

 Africa, by cooperation between the Imperial and the 

 colonial authorities, was put forward by that office 

 long before the date of the similar suggestion by 

 Colonel Morris, referred to on p. i6. 



The whole history of this geodetic work is a 

 curious inversion of the general order. Usually it 

 is the complaint of the map-maker that, whereas it 

 is not difficult to get money from a Government 

 department for the immediate, practical work of map- 

 ping, it is a more laborious task to persuade them 

 of the necessity for a liberal expenditure upon the 

 fundamental geodetic triangulation. In South Africa 

 the exact reverse of this has been the case, and we 

 have tlie anomalous position of a complete triangula- 

 tion system without the resulting maps ; even as vet 

 it is only in the case of the Orange River Colonv 

 and partially in Cape Colony that any of the maps 

 of the country are based upon the positions of the 

 geodetic points. 



Of the technical part of the report the most in- 

 teresting is undoubtedly the account of the base 

 measurements carried out with invar wires hanging 

 freely, at a constant tension, between low tripod 

 supports. Five bases in all, totalling a length of 

 70 miles, were measured. Each was gone over 

 with the wire three times, and the apparent probable 

 error varied from i part in 1,000,000 at the Belfast 

 base, where the staff was inexperienced, down to 

 nearly i part in 7,000,000 in the most favourable 

 case. Sir D. Gill maintains that, with a trained 

 staff, a base can be measured in this way with an 

 actual final uncertainty of less than i part in 1,000,000 

 — say, I inch in 15 miles — a contention apparentlv 

 justified by the figures. The rate of progress, in- 

 cluding the time spent on the wire comparison with 

 the standard bars, averaged 47-^ yards per day, and 

 the cost was high — 15s'- per mile of base. In view 

 of this, and in view of the fact that a limiting error 

 of I part in 1,000,000 implies a much higher degree of 

 precision than that attained by the angular observa- 

 tions, it would seem more practical, for similar work 

 in the future, to make the bases both shorter and 

 less accurate, and, therefore, cheaper and more 

 rapidly executed. This would have the effect of pre- 

 serving that balance between the degrees of precision 

 of the different parts of the work so essential 

 to the economical conduct of a cycle of physical opera- 

 tions. 



The horizontal angles were observed with the 

 lo-inch Repsold theodolite, the probable error of a 

 single angle being found to be o'.-^o with eight 

 changes of zero, or o".3q with four only. It is re- 

 inarked that as these figures closelv coincide with 

 those previously reached in Cape Colony and Natal 

 with the same instrument, thev probably represent 

 the highest possible degree of nrecision attainable 

 under the special climatic conditions and with the 

 instrumental means available. So far as the observ- 

 ing end of each line is concerned this is possibly 

 true, but it is questionable whether the results might 

 not have been improved, with no sacrifice of time or 

 money, if a better pattern of beacon had been em- 

 ployed. The tripod or quadripod beacon, forming 

 from any distant point a double cone, with vertex 

 at the centre, of sufficient height to enable the theo- 

 dolite or heliostat to be centred without disturbing 

 the legs, is an altogether preferable form to the oole 

 beacon actually used. E. H. H. 



