Separation of Bitumen. For the purpose of testing the bitumen in surface mixtures, or in asphalt cement containing considerable mineral matter, it may be necessary to separate the bitumen from the mixture. The following method is given by Mr. Dow: * "The pure bitumen is obtained from an asphalt, or asphaltic cement, by extracting with carbon disulphide and evaporating off fhe solvent. The procedure that I have found to give the best results is as follows: Suffi cient of the asphalt or asphaltic cement to give 3o grams of pure bitumen is placed in a large Erlenmeyer flask. Between 300 and 40o centimeters of carbon disulphide is added, the flask corked and then shaken from time to time until none of the asphalt is seen the sides or bottom, after which the flask is set aside and allowed to stand for 24 hours. The carbon disulphide is then decanted off carefully from the residue into a second flask. The residue is again treated with 200 or 30o cubic centimeters of the solvent and shaken as before. After the solutions in the two flasks have been allowed to subside for 24 hours, the contents are carefully decanted off on to an asbestos filter, passing the contents of the second flask through the filter first. The solvent containing the bitumen is then distilled in a flask until just suffi cient remains to have the contents liquid. It is then poured into a flat evaporating dish and further heated on the steam-bath, stirring from time to time, until the greater part of the carbon disulphide is evaporated. About one-half cubic centimeter of water is next incorporated into the residue of bitumen and the heat ing continued over a burner until all foaming ceases, after which it is kept at 300° F. for 10 minutes. While heating over the burner the bitumen should be stirred constantly with a thermometer and care exer cised that the temperature is kept constant at 300° F. It is doubtful whether in all cases the last traces of carbon disulphide are removed, even by this method, and it is also likely that the bitumen obtained in this way is often slightly harder than that contained in the original asphalt or cement; but its physical prop erties, as far as ductility and susceptibility to change in temperature go, will be relatively the same, and a sufficiently close approximation can be made of the consistency of the bitumen in the original sample to answer all practical purposes. As the removal of the last traces of carbon disulphide is very difficult, and a soft bitumen is liable to be hardened in so doing, I make it a practice, wherever it is possible, to extract the bitumen from an asphalt before it has been soft ened into the paving cement. In this way I find it easier to remove the last traces of solvent from this hard bitumen, and at the same time with relatively less hardening. This bitumen from the asphalt is then fluxed into a paving cement by adding to it an amount of flux equivalent to that used in making the paving cement from the asphalt. It is fortunate that nearly all the asphalts met with in commerce that are not pure bitumen are of a hard nature, so that the above method is applicable in practically all cases. This of course does not apply to bituminous rock, and the only way possible to estimate their quality is by examining the extracted bitumen, which is done as just described. It is well to note here that in cases where the bitumen hardens materially in the removal of the solvent, such a bitumen will be rejected by hardening too much in the heat test."
Tests Required by Specifications. Many tests have been proposed for the control of asphalt paving mix tures, or used in the study of asphalt materials, by various investigators. In general, however, specifi cations used by municipal engineers have depended • upon a contractor's guaranty for the character of the work rather than upon inspection of the materials and workmanship. This has not proven altogether satis factory in many places, and some cities employ expert inspectors and depend upon their own tests for the con trol of the On account of the complex composition of the bitu mens and their peculiar physical properties, tests of the kind ordinarily applied to engineering materials can not be employed in judging asphalt mixtures, and the testing of the materials has been left for the most part to experts employed by the asphalt companies. In only a few instances have specifications looking to the control of the work by test and inspection been em ployed. The specifications used in Kansas City by City Engineer Harper in 1908 contain the following requirements for the asphalt cement: "The asphaltic cement when considered apart from the mineral matter shall have the following characteristics: "It shall be free from water or decomposition products.
"The various hydrocarbons composing it shall be present in homogeneous solution, no oily or granular character being present.
" It must, when tested at 77° F., have a penetration of from 3 to 9 millimeters when tested for 5 seconds with a No. 2 needle weighted with 100 grams, according to the nature of the asphalt and the conditions under which it is employed. It must not be so susceptible to changes of temperature that if at 32° F. it shows a hardness indicated by 1 millimeter penetration, at 151° F. it will not be so soft as to give more than 35 millimeters penetration, using the above method -of testing.
"Twenty grams of it shall not lose more than four (4) per cent in weight upon being maintained at a uniform temperature of 325° F. for seven (7) hours in a cylin drical vessel two and one-half (21) inches in diameter by two (2) inches high.
"Twenty (20) grams of it shall not lose more than eight and one-half (81) per cent upon being maintained at a temperature of 400° F. for seven (7) hours in a cylindrical vessel two and one-half (21) inches in diameter by two (2) inches high.
" It shall be soluble in chemically pure carbon bisul phide at air temperature to the extent of at least ninety five (95) per cent.
" It shall not contain of carbonaceous matter insol uble in chemically pure carbon bisulphide, air tem perature, more than four and one-half (4k) per cent.
" It shall be soluble in 87° Baume petroleum naph tha, air temperature, to the extent of not less than sixty-five (65) per cent and not more than eighty (8o) per cent.
" Its solubility in carbon tetrachloride shall not be more than one and one-half (II) per cent less than its solubility in carbon bisulphide — both tests being made at air temperature.
" It shall show of fixed carbon not more than fifteen (t5) per cent.
" It shall show a flashing point (New York State Closed Oil Tester) of more than 350° F.
"It shall not contain more than three (3) per cent of paraffin scale, the Holde method of determining paraffin scale being used.
" The bitumen entering into the composition shall have been in use in the street paving industry under conditions similar to those contemplated in this con tract, at least 4 years prior to the letting of this con tract."