May 31, 1918] 



SCIENCE 



543 



There is, however, a serious difficulty with 

 the data from both sources. This difficulty 

 lies in the fact that the average number of 

 daughters per mother is extremely small. The 

 average number of daughters per mother in 

 Pearl's experiment was 2.85, while in mine it 

 varies from 2.6 in the early years to 6.75 in 

 later years. Because of the small size of the 

 families it is possible to fit any family into a 

 place in either scheme, since the ratios ex- 

 pected for the various matings differ only 

 slightly from one another. In spite of the 

 doubt raised as to the mode of inheritance of 

 winter egg production it is clear that this 

 character is inherited, for high and low fe- 

 cundity lines are readily established by suit- 

 able matings along family lines. 



h. d. goodale 

 Massachusetts Agricultxjral 

 Experiment Station 



THE AMERICAN ASSOCIATION FOR 



THE ADVANCEMENT OF 



SCIENCE 



SECTION D— ENGINEERING 



The first session was held on the morning of 

 Friday, December 28, in Thaw Hall, University of 

 Pittsburgh, Vice-president Dr. Henry S. Drinker 

 in the chair, with an attendance of about thirty. 

 It was announced that the Sectional Committee 

 had recommended for election to the General Com- 

 mittee, for the office of vice-president, Dr. Ira N. 

 HoUis, of Worcester, and for the office of secre- 

 tary. Dr. Frederic L. Bishop, University of Pitts- 

 burgh. The following officers were elected by the 

 Section : 



Member of Council — Dr. George F. Swain, of 

 Harvard University and the Massachusetts Insti- 

 tute of Technology. 



Member of General Committee — Charles Henry 

 Davis, of Cambridge, Mass. 



Me7nber of Sectional Committee — William Bowie, 

 of Washington, D. C. 



The program of the session was as follows: 



Railroad track, its defects and abuses, and 

 their amelioration : G. H. Barbour. Histor- 

 ical: The age of the drawn and that of the 

 driven wheel; ancient English tramways; the 

 institution of that distinctively American prin- 

 ciple now governing the construction of railroad 



track as now practised by all the steamroads in the 

 world, wherein the equijinient upon its track con- 

 stitutes a flexible superstructure upon an elastic 

 roadbed. Defects: Weak rails; narrow railheads; 

 excessive deflections; joints. Abuses: Worn and 

 ragged wheels; improper wheel spacing; dynamic 

 aug^ments; lateral thrusts. Ameloriation : In- 

 creased bearing on ballast ; decreased depth of bal- 

 last; augmentation of rail; increased lateral 

 strength; broader head; more frequent lateral fas- 

 tenings; maintaining height of rail at the mini- 

 mum. 



The scientific principles of building codes: J. A. 

 Ferguson. A good building code occupies a very 

 responsible position among the vital issues of mu- 

 nicipal welfare. Properly planned, a building code 

 should insure safety to life, limb, health and prop- 

 erty, and should function to minimize loss or in- 

 jury to either. Progress in the arts has introduced 

 many new factors in the occupation of buildings, 

 which necessitate the scientific handling and 

 classification of the requirements and progress in 

 building has made it possible to classify the vari- 

 ous forms of building construction into distinct 

 groups. The same progress has made it possible to 

 classify occupancies and construction of buildings 

 and to specify the minimum allowable construc- 

 tion for the various occupancies. This it is now 

 proposed to do in one notable case for the city of 

 Pittsburgh, Pa. Other phases of this subject are 

 susceptible of scientific definition, and in order to 

 properly regulate buildings it is becoming increas- 

 ingly necessary to classify and define in a scientific 

 manner all subjects. The paper gives typical ar- 

 rangement for a code and explains the reasoning 

 upon which it is based as well as for the classifi- 

 cation of other regulatory pro\"isions in a good 

 building code. 



Belative efficiency of different methods of re- 

 pairing bituminous macadam and bituminous con- 

 crete pavements: George H. Biles. The bitu- 

 minous macadam and bituminous concrete pave- 

 ments in their various stages of disrepair offer ex- 

 cellent opportunity to the highway engineer for 

 study and experiment. The methods of repairs to 

 pavements of these types have advanced to such a 

 degree in recent years that there are innumerable 

 instances where pavements have been reclaimed 

 by scientific analysis of the causes of deteriora- 

 tion and by efficient application of the principles 

 of repair applicable to each case. Central bi- 

 tuminous mixing plants are advisable where the 

 amount of yardage and its accessibility warrant 

 as in cases of municipalities. In most other cases, 



