252 FOREST TREES OF NORTH AMERICA. 



separated into resiuons and uon-resiuous classes. Tbe specific gravity gives a direct means of comparing heat 

 values of equal volumes of wood of different resinous and nou resinous species. In burning wood, however, various 

 circumstances affect its value; few fire-places are constructed to fully utilize the fuel value of resinous wood, and 

 carbon escapes unconsumed in the form of smoke. Pine, therefore, which, although capable of yielding more heat 

 than oak or hickory, may in i)raetice yield considerably less, the pine losing both carbon and hydrogen in the form 

 of smoke, while hickory or oak, burning with a smokeless flame, is practically entirely consumed. The ash in a 

 wood, being non-combustible, influences its fuel value in proportion to its amount. The state of dryness of wood 

 also has much influence uixm its fuel value, though to a less degree than is generally supposed. The water in 

 green wood prevents its rapid combustion, evaporation reducing the tcmperiiture below the point of ignitioD, 

 Green wood may often contain as much as 50 per cent, of water, and this water must evaporate during combustion ; 

 but as half a kilogram of ordinary wood will give 2,000 units of heat, while half a kilogram of water requires only 

 268.5 units to evaporate it, 1731.5 units remain available for generating heat in wood containing even a maximum 

 amount of water. In cases where tlie i>ressure was perpendicular to the grain of the wood it was applied on the 

 side of the specimen nearest to the henrt of the tree. 



A factor in the general value of wood as fuel is the ease with which it can be seasoned ; beech, for example, a 

 very dense wood of high fuel value when dried, is generally considered of little value as fuel, on account of the 

 rapidity with which it decays when cut and the consequent loss of carbon by decomposition. 



THE STRENGTH OF WOOD. 



The specimens tested for the purpose of determining the strength of the wood produced by the different trees 

 of the United States were cut, with few exceptions, before Msych, 1881, and were slowly and carefully seasoned. 



Those used in determining the resistance to transverse strain were made 4 centimeters square and long enough 

 to give the necessary bearing upon the supports. These were shod with flat iron plates, slightly rounded on the 

 edges and were set exactly 1 meter apart; they remained perfectly rigid under the pressure applied. Each specimen 

 was weighed, measured, and its specific gravity calculated before it was tested. The result thus obtained represents 

 the specific gravity of the air-dried wood. 



To eliminate the action of their weight the specimens were placed upright, and hydraulic pressure was applied 

 by means of an iron rod 12 millimeters in radius, acting midway between the supports, the deflections being read 

 at this point. 



The direction of the grain of the wood is shown by diagrams in the table (Table III), the pressure acting upon 

 it horizontally from the left. 



The pressure was applied slowly and uniformly, a reading of tbe deflections being taken for every 50 kilograms. 



When a load of 200 kilograms had been applied it was removed and the set read. Pressure was again applied in 



the same way, and the readings of deflections were resumed when 200 kilograms was again reached. 



P P 

 The formula used in calculating the coefficient of elasticity was E = . . ,3 ; I, b, d, being taken in millimeters; 



3 P J 

 that of the modulus of rupture, ^=~2~h~^) h ^j '^ being in centimeters, P, in both formulas, in kilograms. 



A few experiments were also made in the same manner, for purposes of comparison, to determine the transverse 

 strength of specimens 1 meter long between the bearings and 8 centimeters square (Table IV). 



The specimens tested by longitudinal compression were 4 centimeters square and 32 centimeters (8 diameters) 

 long. They were jjlaced between the platforms of the machine, and pressure was gradually applied until they 

 foiled. The figures given represent the number of kilograms required to cause failure. 



The specimens tested under pressure applied perpendicularly to the fibers were 4 centimeters square and 16 

 centimeters long. They were placed upon the platform of the machine and indented with an iron punch 4 

 centimeters square on its face, covering the entire width of the specimen and one-quarter of its length at the " 



center. In this series of experiments the direction of the annual rings was noted, horizontal pressure being also 

 ai)plied from the left. Readings were taken of the pressure necessary to produce each successive indentation of 

 0.254 up to 2.54 millimeters, and in the case of specimens which did not fail with this pressure a further test was 

 made of the weight required to produce indentations of 3.81 and 5.08. The remarks (Table V) upon the behavior 

 of the wood of the different species under compression were furnished by Mr. James E. Howard, in charge of the 

 testing machine. 



COMPARATIVE VALUES. 



In tbe following table the number standing opposite each species represents its relative value in the column in 

 which it appears. 



This table is purely an arbitrary one, since the introduction of one or more species would of course change the 

 value of all species standing lower in value, or results based on an examination of a larger number of specimens 

 of any species may change the relative numbers in regard to it very considerably. In other words, any twenty or 

 thirty si)ecies bearing consecutive numbers may change places with each other. This arises partly from the want S 



of uniformity of the wood of any species, and partly from the fact that where so many determinations fall between f 



comparatively narrow limits the mere order of sequence must be largely accidental. f 



