July 25, 1901] 



NATURE 



Z^l 



cent, on two inches, correlated, however, with a maximum stress 

 of only I9'2 tons per sq. in. 



.\ long series of observations revealed the curious fact that 

 no correlation exists between the densities and mechanical 

 properties of steel castings. 



The mechanical influence of annealing at 950° C. and slowly 

 cooling from that temperature is, in the case of iron containing 

 about 0'4 per cent, of carbon, exemplified by the following 



Max. stress EIongati( 



Redu 



The micrographic effect of drastic annealing is shown in the 

 drawing, cc 2, which has been reduced from a 6-inch circle 

 magnified 315 diameters. 



The constituents of the steel as cast present an irregular 

 trellis-like pattern of pale ferrite or iron and dark pearlite or 

 steel. In addition, small castings or small parta of large cast- 

 ings, present curious brown-etching lines of a sulphur alloy. 



Fig. I.— Casting CC;. (Reduced from s 



which arrange themselves almost exclusively along the ferrite, 

 forming lines of dangerous weakness. 



After annealing, the long lines of cleavage between the con- 

 stituents are broken up, large patches of laminated pearlite and 

 allotrimorphic crystals of iron being formed. At the same time, 

 the dangerous lines of the attenuated sulphur alloy are destroyed, 

 segregating into isolated patches. All the above features are 

 well indicated in the micrograph figured. Steel castings often 

 present mechanical discrepancies very difficult to explam, e.i^. 

 elongations per cent, and bending angles form measures of duc- 

 tility which might be expected to be proportional to each other, 

 but often such is not the case. 



.\ remarkable feature of the results recorded is the fact that 

 some of the castings built up of very large crystals have pre- 

 sented great ductility, whilst castings with that minutely crystal- 

 line structure usually supposed to give the best mechanical results 

 have given unsatisfactory tests. 



Considering the fact that such castings were of identical 

 chemical composition and had been subjected to similar thermal 

 treatment, it is difficult to avoid the conclusion that the initial 

 temperature of the steel on casting may exert a permanent 

 mechanical influence, and that consequently the operation of 

 annealing is not thoroughly effective. 



SCIENTIFIC WORK IN EGYPT. 

 CEVER.\L matters of scientific interest are described in Lord 

 "-^ Cromer's report upon the finances, administration and 

 condition of Egypt and the .Soudan. Particular attention is 

 directed to Sir William Garstin's memorandum on irrigation, 

 and to the value of the work of hydraulic engineers in Egypt. 

 Sir William Garstin brings forward observations showing a con- 

 tinuous fall of level of Lake Victoria Nyanza. His remarks 

 upon this fall and the rainfall observations available are here 

 reprinted. 



" Lake Victoria Nyanza Rain Ganges. — Information regarding 

 the daily rainfall and the lake levels is now regularly received 

 in Egypt from Uganda. The first series of registers, previous 

 to the rebellion, dates from June i, 1S96, to July 31, 1897. 

 The second series commences on September i, 1898, and has 

 been carried on to October 31, 1900. Unfortunately, it is 

 impossible to connect the two series, as the gauges erected 

 subsequent to the rebellion are not the same as those which 

 formerly existed. We have, therefore, only twenty-six months 

 of observations upon which to base possible theories. This is 

 not sufficient, more especially as we are still ignorant of many 

 important factors bearing upon the relations of the levels and 

 rainfall of the equatorial lakes with the Nile supply. Informa- 

 tion regarding the most important point of all, viz. the Albert 

 Nyanza Lake, is still entirely wanting. This is much to be 

 regretted, as this lake, which drains an enormous catchment 

 area, and through the northern end of which passes the water 

 coming from the Victoria Lake, is probably the most important 

 of the reservoirs which leed the White Nile. It is to be hoped 

 that early measures may be taken to erect gauges upon the 

 Albert Lake and to observe them regularly. The records thus 

 obtained would be invaluable to Egypt. 



" The following facts, elicited from the rather meagre informa- 

 tion at our disposal, may perhaps be of general interest : — 



" Two observing stations now exist on the Victoria Lake, one 

 on either side, viz. at Port Alice or Entebbe and at Port 

 Victoria or Ugoive. Unfortunately, the rainfall register for 

 Entebbe only commences from the month of April 1900, so a 

 comparison between the two for the whole year is impossible. 



"The Ugowe rain gauge record shows that 46 -28 inches of rain 

 fell in the twelve months ending with October 1900, that there 

 were 131 rainy days, and that the storms invariably took place 

 either in the afternoon or at night. Further, that February 

 was the wettest month of the year and July the driest, 6-45 

 inches being registered for the former and 1-56 inchss for the 

 latter. It would also appear that the period of the heaviest 

 rainfall is from November to May. The total rainfall for the 

 six months in question was 30-73 inches, the remaining six 

 months being responsible for only 15-55 inches. 



"The rainfall at Entebbe, on the opposite side of the lake, 

 was, so far as the records go, considerably heavier than at 

 Ugowe. The total fall between .\pril and November 1900 was 

 30-39 inches at the former place, as against 2059 inches at the 

 latter. As regards the rise and fall of the water surface of the 

 lake it is possible to make a comparison, as registers have been 

 supplied regularly from both stations. . . . 



" The records, as far as they go, seem to prove that the lake is 

 always at its lowest in October, i.e. at the end of the dry 

 season, and at its highest in December or January. They also 

 show that the lake level has been steadily falling for the last 

 three years, thus : — 



Port Alice 

 (Entebbe). 

 Ft. in. 

 Lake levels on October I, iS 



Port Victor! 

 (Ugowe). 



1899 

 1900 



"It is possible that this fall in the levels may be, in some 

 measure, due to degradation of the bed of the river in the 

 channel whence it issues from the lake ; but it seems to be far 

 more probable that it has been caused by a failure of the rain- 

 fall over an immense area. All reports go to show, and all 

 travellers who have visited these regions relate, that a severe 

 drought has prevailed over a large portion of Central .Africa 

 during the last two years. This drought has extended as far 

 north as the Egyptian Soudan. 



" .Although it may seem likely that the Nile flood of 1901 will 

 be a poor one if these figures are correct, it would be misleading 

 to attempt to draw any definite conclusions from them. The 



NO. 1656, VOL. 64] 



