April 1952
INDUSTRIAL A N D ENGINEERING CHEMISTRY ACKNOWLEDGMENT
The authors are indebted to M. J. Den Herder for conducting the catalytic cracking studies. LITERATURE CITED
(1) “A.S.T.M. Standards on Petroleum Products and Lubricants,” p. 77, Method D 90-47T, Philadelphia, Am. SOC.Testing Materials, 1949. (2) Ibid., p. 139, Method D 129-49. (3) Ibid., p. 355, Method D 189-46. (4) Ibid., p. 706, Method D 86-46. (5) Barieau, R. E., and Barusch, M. R., Petroleum Division Preprints, p. 191, 115th Meeting AM, CHEM.SOC.,San Francisco, April 1949. (6) Evering, B. L., and d’Ouville, E. L., J . Am. Chem. Soc., 71, 440 (1949). (7) Grosse, A . V., and Wackher, R. C., IND.ENG.CHEM.,ANAL. ED., 11, 614 (1939).
879
(8) Hughes, E. C., Scovill, W. E., Whitacol, C. H., Faris, R. E., Bartleson, J. D., and Darling, S. M., IND.ENG.CHEM., 43, 750 (1951). (9) IND.ENG.CKEM.,41, 2681-740 (1949). (10) Kalichevsky, V. A., and Stagner, B. A., “Chemical Refining of Petroleum,” ACS Monograph 63, New York, Reinhold Publishing Corp., 1942. (11) Lien, A. P., McCaulay, D. A., and Evering, B. L., IND.ENG. CHEM.,4 1 , 2 6 9 8 (1949). (12) Meadow, J. R., and White, T . A., Zbid., 4 2 , 9 2 5 (1950). (13) Scafe, E. T . , Petroleum Refiner, 25, 413 (1946). (14) Schneider, K. W., and Feichtinger, H., Angew. Chem., B20, 1216 (1948). (15) Schneider, K. W., and Gottschall, H., ErdoZ u. Rohle, 1 , 7 4 (1948). (16) Tamele, M., and Ryland, L. B., IND. ENG.CHEM.,ANAL.ED.,8. 16 (1936). RECEIVED for review November 1 , 1960. ACCEPTED November 8, 1951. Presented before the Division of Petroleum Chemistry at the 119th Meeting of the A n f E R I C A m CREMICAL SOCIETY,Boston, Mrtss.
Rubber Recovery from Freshly Harvested Guayule -
K.
J
w. TAYLOR AND R. L. CHUBB
U . S. Natural Rubber Research Station, Salinas, Calv.
c-$-uui:
4
rubber from guayule has been produced commercially
in Mexico and the United States for nearly 50 years (11). Much of the production in both countries has been carried out by the Intercontinental Rubber Co. or its progenitors At the beginning of World War I1 all of the holdings of the Intercontinental Rubber Co. in the United States were purchased by the Government, and the Emergency Rubber Project was created to produce natural rubber from guayule and other plants which could be domestically grown (6). The Emergency Rubber Project was ordered liquidated shortly after V-J day and the research on guayule processing came to a halt. Natural rubber is still essential (6) and has been declared a strategic and critical material. About 25% of the rubber requirements of the United States must be natural rubber, and over 90% of the world’s natural rubber lies thousands of miles away in politically and geographically vulnerable areas. In August 1947 the U. S. Natural Rubber Research Station was set up to continue research on natural rubbers which could be domestically produced for use in the event of a national emergency, or which exhibited potentialities as an additional economic agricultural crop. Current process research on guayule looks to the development of new and improvement of old production methods. It is the purpose of this paper to report the development of one such improvement HISTORICAL
Guayule (Parhenium argentatum, Gray) is a woody, semidesert shrub not unlike sagebrush in appearance, that is native t o northern Mexico and to the Big Bend region of Texas in the United States (7). It has long been of interest as a producer of rubber, and selected strains have been known to produce as much as 2001, of their dry weight as rubber hydrocarbon. Recent hybridization of guayule with other species of the same genus holds promise of increased rubber production per acre (14). Rubber in guayule, as in other rubber-bearing plants, is present as latex. It is usually recovered in solid form, because the latex is borne in discrete cells rather than in latex ducts that can be
tapped. Rubber in h e x form can be recovered from guayule, but this has not been demonstrated to be economically feasible. Ever since the inception of the commercial milling of guayule for rubber the consensus has been that provision for the storage of shrub prior to milling was necessary. The varied types of storage commercially used in the past consisted of prolonged field curing; holding baled foliate material in bodegas (bodega is a term inherited from the early Mexican guayule industry and refers to storage houses where the shrub was “cured” before milling) or out of doors under a variety of conditions; holding foliate chopped shrub in bins; or holding baled defoliated shrub under rather extreme drying conditions. The reasons advanced to justify the storage of shrub were almost as numerous aa the kinds of storage employed. I n some instances it was necessary to provide shrub storage so that mill operation could be continuous in spite of inclement weather. I n other instances it wm regarded as advantageous to field cure shrub, which after all is, in part, another method of storage, to reduce through moisture loss the weight of the material to be transported to the mill. This became particularly important when the hauling distance to the mill was great. I n all instances, it was contended that storage was necessary to bring about proper “conditioning” (11) of the shrub so that the maximum amount of crude rubber could be recovered in the milling process. In other worde, i t was believed necessary to coagulate all the latex by reducing the moisture content of the rubber-bearing cells before processing was begun. A great amount of study has been devoted to shrub storage. Such studies have included compressing the shrub, either whole or chopped, into compact blocks; holding it in cold storage; holding it under aerobic (retting) conditions ( 2 ) ; or holding i t under anaerobic (ensilage) conditions. Abortive attempts were even made to hold foliate chopped shrub in storage bins where the normal atmosphere in the shrub mass was replaced with stack gases or other presumably inert materials. I n addition to these studies, investigative work was done along the lines of storage used in commercial processes.
I
880
INDUSTRIAL AND ENGINEERING CHEMISTRY
With the beginning of mill operation by the Emergency Rubber Project there evolved a method of shrub condit,ioning which involved both field curing, commonly called sunning, and bodega storage. Out-of-doors stockpiling supplemented bodega storage when t.he bodegas were full and it was feared inclement weather would cause a cessation in harvest activities. As a general rule, field curing consisted of throwing two field rows together in a windrow and leaving them for 5 days or more. The shrub was baled in the field and transported t'o the mills where it was held in either bodegas or stockpiles until it was judged to be ready for milling, The period of holding varied from 10 to 45 days. It is apparent that with such variation in the mode of storage there was bound to be a decided variation in the characteristics of the raw material entering the milling circuit. However, a t the conclusion of the Emergency Rubber Project essentially this same manner of shrub handling persisted. The current practice in the commercial production of guayule rubber in Mexico is similar except that there is even more variation in the field cure period, Current process and development research on guayule is based on the premise that the shrub and its constituents possess the greatest uniformity as a stand of living pIant,s in the field when they have been afforded reasonably uniform culture throughout their lives. When the plants are removed from their field environment there begins a myriad of complex changes which vary in their rate and extent directly with the-intensity of the surrounding physical and biologic factors. The authors, while reviewing past experimental work ( 2 , 8) preliminary t o the development 'of an over-all plan for process research and development, noted ,occasional references to the milling of freshly harvested shrub without any conditioning. I n addition, the work carried out in the development of pilot plant cont,rol (12) demonstrated that :recoveries of about 90% of the rubber hydrocarbon could be ohtained from fresh shrub fairly easily. These facts led to the consideration that complete coagulation OF the latex normally brought about by shrub conditioning could be achieved by a combination of parboiling and mechanical treatment of the fresh shrub. This should enable maximum recoveries and would obviate the hazards engendered by conditioning. This view was enhanced by certain indications that conditioning might be deleterious. Furthermore, other aspects of milling freshly harvested shrub appeared economically attractive. Accordingly, a series of experiments was performed to determine the maximum rubber hydrocarbon recovery obtainable from fresh shrub and to compare the crude rubber produced to crudes obtained from shrub conditioned in the conventional manner. EXPERI1\IE>'CAL WORK
Vol. 44, No. 4
the ensilage cutter and the rotary fly-knife cutter. The cut shrub was divided into 8 replicates and each was sampled for the determination of rubber and resin (IO). Four replicates were used as controls and each was crushed twice, then handled in the standard control manner (18). The other four replicates were handled in the same manner except that each replicate was crushed six times. The following day, the eight replicates were milled in a battery of four 27-inch batch mills. The experiment was so designed that each batch mill received one mill charge from each of the two different crusher treatments. The milling order was designed so that each treatment was represented in the first four and second four millings to distribute any differences due to time lag ( l a ) . Milling was carried out in accordance with the control procedure. When milling was completed the slurry was discharged into a flotation tank and the mills were thoroughly rinsed to remove any rubber or other plant material which was entrapped. The floating rubber and cork were skimmed off into holding cans, and representative samples of bagasse and effluent were taken for chemical analysis of rubber hydrocarbon, resins, and total solids. The rubber from each replicate was paila decorked (S), scrub milled, and dried. A 50-gram sample was removed and analyzed for rubber hydrocarbon, resins, insolubles, and moisture. While the additional crushing did give increased rubber hydrocarbon yields (as will be discussed later), there was some reason to believe that a hammer mill might prove more effective in rupturing rubber-bearing cells. Hence, an experiment was designed which would afford some measure of the comparative effectiveness of crushing and hammer milling in various combinations. The procedure for the second experiment involving the varied crusher and hammermill treatments was identical with that just described except for the following. After the shrub was cut it was divided into eight replicates as before. The eight replicates were then divided into four groups of two replicates each. The first group was crushed twice and hammer milled once. The second was crushed twice and hammer milled twice. The third was crushed four times and hammer milled once. The fourth and last group was crushed four times and hammer milled twice. Samples for chemical analysis and moisture determination were taken by passing quart jars intermittently through the discharge stream from the hammer mill to eliminate the drying and coagulating that riffling down in a binary sample divider might effect. All milling and subsequent steps were carried out as wm done in the first experiment. COAGUL~TION nY STORAGE. An experiment was carried out in which one mass of shrub was treated to induce coagulation by mechanical means, and another had coagulation brought about by field curing and storage of shrub in a manner comparable to that used in the commercial production of guayule rubber. Comparisons were then drawn as to the quantity and quality of the crude rubbers recovered.
O F ADDITIONAL CRUSHING ON RUBBER I. EFFECT HYDROCARBON^ DISTRIBUTION IN MILLENDPRODUCTS
TABLE
Throughout the course of these PUIATE~IAmAND EQUIPMENT. Adiusted Rkbber experiments, 5-year-old dryland guayule, variety 593, from a Rubber IIydrocarbonb, % . Hydrofield approximately 5 miles south of Salinas, Calif., wa8 used. RecovLFst Total carbonc Recovefed Lpst The pilot plant equipment included that required for pilot plant In accountered in crude in bagasse in effluent ability Crude control millings ( l a )and a manual baling press. 85.3 5.7 103.1 90.1 7.3 PROCEDURE. COAGULATION BY MECHANICAL TREATMENT. Crushed 2 times 93 . a 1 4 106.6 98.5 5.6, Crushed 6 times Two different experiments were carried out to determine the 2.5 8.0 4.3 8.4 1 . 1 Mean difference Difference required effect of combined parboiling and mechanical treatment on the for significance a t 0 6 1.8 5.4 Not 1% level coagulation of the latex and subsequent recovery of rubber. One significant of these dealt with an increase in the amount of crushing given a The consistency of rubber hydrocarbon accountability by chemical the shrub; the other compared the effect of varied crushing and anal sis of the mil1,products that indicetes mpre rubber. than can be sh0p.n b y cxemical analysis of the shrub has given rise to considerable specu!ation hammermill treatments. For the sake of clarity, the procedural in the past ( I f ) . It must be inferred that the best criterion for estimation of rubber in the plant is by analysis of the mlll products-crude rubber, badetails for the two experiments will be discussed separately. gasse and effluent. The success of chemical analysis of the shrub by pres-
An area in the field was selected where conditions of growth
and culture were uniform. This area was rogued and 250 of the remaining plants were harvested by hand, undercutting to .a depth of about 6 inches. The harvested plants were transported to the pilot plant without delay. There they were parboiled ( I ) for 15 minutes a t an average temperature of 200" F. The parboiled shrub was defoliated in the screened trommel and passed consecutively through
entlv'apuroved methods is based primarily upon extractmn of the rubber from the' plant material with suitable sqlvents. To say, then., t h a t batch mill balances are the best criteria for estimation of rubber is simply t o say t h a t the mill products are more completely and readily extracted by the solvent in rubber analysis than is shrub material. b Per cent of total rubber hydrocarbon originally present in the plant as determined by chemical analvsis. 0 Adjusted recovery of rubber hydrocarbon based on total accountability at bv chemical analvsis of the mill end uroducts. Values given are arrived ... means of fo& replicate mill-charges.
April 1952
INDUSTRIAL AND ENGINEERING CHEMISTRY
I n general, the procedure consisted of milling defoliated lush shrub, and shrub which had been dug, field cured 7 days, baled and stored, then parboiled and defoliated as the bales were removed for milling. The defoliated lush shrub constituted the control. The field-cured shrub followed the conventional commercial practice except that all shrub was field cured for the same specific time, After storage, millings were made each week for 8 weeks to establish trends. The shrub used was from the same field and of the same age class as that used in the mechanical coagulation studies. An area of uniform shrub was chosen and the plants to be used for lush milling and for storage were designated by predetermined randomization. All plants were dug a t the same time and those used as controls were processed a t once in %mill charges according t o the standard control milling procedure. The dry weight of these lants was determined and used to.establish the original dry weig1t of the plants placed in storage. At the conclusion of the field cure the plants were baled, an equal number of plants being incorporated in each bale. Each bale was wrapped in cheesecloth t o preclude loss of plant material. All bales were placed in storage under conditions comparable t o those commercially used. Records of dry weight changes were maintained throughout the experiment. A t the end of the first week of storage and for 7 weeks thereafter, four bales were removed from storage in a predetermined random order and, after being weighed, were defoliated and four 27-inch batch pebble mill charges were processed. All shrub preparation, milling, and sampling were carried out in the same manner as for the controls. DISCUSSION
RESULTSO F MECHANICAL COAGULATION. In the first experiment, where the effect of additional crushing was compared t o a standardized control method of milling, it was apparent that yields of rubber hydrocarbon recovered in the crude rubber were markedly increased. Table I illustrates the effect of crushing on rubber hydrocarbon recovery. The data when analyzed by Fisher’s F test (9) indicate that the odds are less than one in a hundred that‘ the increased rubber hydrocarbon recovery is due t o chance. Added crushing therefore can be judged effective. I n Table I it is shown that the material crushed twice shows a total rubber hydrocarbon accountability of 103.1$Z.0, However, 103.1% is obviously misleading, since the remaining millings from the same lot of $rub which were crushed six times show a total accountability of 105.6’%-a statistically highly significant difference. This figure provides the best estimate of the rubber hydrocarbon actually present in the shrub. On that basis, the rubber hydrocarbon recovered from the shrub crushed twice was only 85.3% of that available. It will be readily seen that the total accountability fluctuates with variation in treatment, and therefore accurate interpretation of data on that basis is difficult and may sometimes be misleading. The amount of rubber as estimated by chemical analysis is consistent, though apparently low and for the sake of simplicity is generally used as the basis of interpreting treatment differences. It is possible by further analysis of the experimental data to determine just how crushing increases the rubber hydrocarbon yield, It might be argued that repeated passage through the crusher rolls had so thoroughly comminuted the shrub that the rubber was more readily and completely recovered from the plant tissues in the milling time of 60 minutw. However, this is not entirely the case. It appears that there is a lower rubber hydrocarbon content in the bagasse from the shrub that had been subjected t o six crushings, though the difference cannot be shown as statistically significant in this experiment which consisted of only four replicates of each treatment. Even if this difference in rubber hydrocarbon losses in bagasse is real, it accounts for only a portion of the difference between yields from the two treatments. T h e increase in the amount of rubber recovered is in greater part due t o the decrease in dispersed latex rubber in the effluent liquor, as shown in Table I. The odds are far in excess of 100 to 1 that the reduction of rubber in the effluent liquor can be ascribed t o t h e additional crushing rather than to chance. This means that more of the latex is coagulated by mechanical means. This is verified by the observation that latex collects as a white cream on
881
the top of either whole slurry or effluent samples from control millings (crushed twice) which have been held for some time. Rubber recoveries in the second experiment demonstrate that the kind of comminution given the shrub before milling has a direct bearing upon the amount of rubber hydrocarbon recovered in the crude rubber. Table I1 shows the mean values of rubber hydrocarbon recovery for each of the treatments.
TABLE 11. RUBBERHYDROCARBON RECOVERED IN CRUDE RUBBERFROM SHRUBCOMMINUTED IN VARIOUS WAYS Hydrocarbon Shrub Treatment Recovery4 Crushed twice, hammer milled once 90.8 Crushed twice, hammer milled twin? 96.8 97.0 Crushed four times, hammer milled once 100.6 Crushed four times, hammer milled twice a Per aent of total rubber hydrocarbon originally present in the plant material as determined by chemical analysis. Values shown are means of t w o treatments. Difference between means required for significance at 5 % level = 6.2%.
It is apparent that one extra hammer milling is as effective as two extra crushings, and almost as effective as two extra crushings and one extra hammer milling. Table I11 illustrates the rubber left in the bagasse and in the effluent from the various treatments.
TABLE111. RESIDUAL RUBBERHYDROCARBON I N BAGASSE AND EFFLUENT FROM SHRUBGIVENVARIEDCOMMINUTION TREATMENTS BEFORE MILLING Hydrocarbon5 In In Shrub Treatment bagasse eauent Crushed twice hammer milled once 7.7 6.2 Crushed tvioe’ hammer milled twice 6.0 3.2 Crushed four Limes, hammer milled once 6 .O 3.4 Crushed four times, hammer milled twice 6.7 3.3 Per cent of total rubber hydrocarbon ori inally present in the plant material as determined b y chemical analysis. &slues shown are means of two replicates.
As in the crushing experiment, most of the increase of rubber hydrocarbon in the crude is obtained from additional coagulation of the latex. While the differences in rubber hydrocarbon loss in both bagasse and effluent are not significantly different, the mean results are in close agreement with those obtained in the experiment where the shrub was crushed two and six times. Table IV gives the mean values for the const,ituents of the crude rubber from the first experiment.
TABLE IV.
-
CONSTITUENTS OF CRUDE RUBBERFROM SHRUB CRUSHED Two AND SIX TIMES^ Rubber Hydrocarbon,
Shrub Treatment % Crushed twice 71.7 Crushed six times 72.8 Observed F values 33.28 a Values given are means of four replicates. means are highly significant. F = 18.74 at 0.01 and 6 degrees of freedom.
Resins,
InsolubIes, % 21.5 7.5 20.4 6.6 83.16 21.81 The differences between level of probability for 1
%
It will be observed that the chemical characteristics indicate an improvement in the quality of the crude produced from shrub given additional crushing. Table V, of constituents and viscosities, further illustrates that mechanical treatment of the shrub prior to milling does not degrade the crude rubber. The viscosities are measured in two waysby the Mooney viscometer (13) and by the Ostwald-CannonFenske viscosimeter. The latter values are expressed as molecular weights (4). While there are not enough replicates to
Vol. 44, No. 4
INDUSTRIAL AND ENGINEERING CHEMISTRY
882 TABLE IT. COIiSTITUEKTS A S D FROM
CRUDE RUBBERS SHRUBP R ~ P A RIK ED VARIOVS WAYSFOR MILLISG
2 1 2
4
4
OF
(Means of two reulioatos) Rubber Hydrolnsolucarbon, Resins, bles,
Shrub Preparation Times Tinies hammer criished milled 2 1
2
\’ISCOaITIES
yo
%
%
69.3 68.8 68.7
22.9 22.8
6.6 6.8
70.4
21.7
6.7
22.1
6.6
Mooney
hlolecu-
ML, lar 212’,B., Weight 5 Min. ( X 1000) 46.5 126 48.5 134 47.0 131 47.0 132
permit the determination of a statistical difference, if any, the evidence amply demonstrates that there is no deterioration of the rubber hydrocarbon; nor does additional premilling treatment lead to increased amounts of the undesirable constituents, resins and insolubles, in the crude,
carbon recovered, but of more significance, the character of the crude rubber changed, See Table VI. There is a very high degree of correlation between the duration of storage and the insoluble content of the crude rubber, the molecular weights, and the Rlooney viscosities. The reduction in the resin content is not great enough to he statistically significant, nor does it have any discernible effect on the other physical properties of the rubber. For every 5 weeks of storage the insoluble content increased about 1%. There was a deterioration of the crude rubber, as measured by the molecular weight and Mooney viscosity. This deterioration occurred at the rate of approximately 6000 units molecular weight and l i / 4 Mooney points per week. The drop in molecular weight and Mooney value is strongly indicative of degradation of the rubber hydrocarbon molecules, possibly by oxidation. Whether this degradation results in poor keeping qualities of the crude as well as in inferior physical properties of the vulcanizates remains t o be studied. Since the inception of these experiments many millings from freshly harvested shrub have been made during all seasons of the year, and from shrub of varied age classes and fields. I n all instances the rubber hydrocarbon was consistently high. CONCLUSIOYS
VI
.-
i
l
I
l l ! I l
2 9
1
2
Freshly harvested guayule shrub can be processed successfully for the recovery of rubber without the addition of coagulant chemicals. Coagulation of the latex in fresh guayule may be brought about by parboiling and mechanical treatment, though the most efficient method for doing so remains to be developed. Mill yields of rubber from freshly harvested shrub are generally high, though the method of shrub preparation influences rubber hydrocarbon recovery. Shrub conditioned and stored in the manner usually employed in commercial operations gives decreasing rubber hydrocarbon yields Kith increased periods of storage. Rubber from shrub handled in this manner exhibits progressive deterioration.
3 WEEXS
4
5
I 6
7
8
9
FROM H A R V E S T
Figure 1. Effect of Shrub Storage before Milling on R u b b e r Hydrocarbon Yields and Losses
RESCLTS OF COAGULATIOX BY STORLGE.The results of this experiment clearly point t o the undeairability of the conventional practices for’conditioning guayule shrub. Taking all changes of weight into consideration, the amount of rubber hydrocarbon recovered by milling decreased with an increase in the length of the storage period. The extent of the decrease is indicated b y the regression plotted in Figure 1. As storage continues, the losses of rubber hydrocarbon in the effluent and bagasse increase. This confirms practical observations made by the senior author in two years of commercirtl factory operation. The losses sustained in this carefully controlled experiment do not approximate those occurring in a factory, because in commercial operations it is not practicable t o r e h e the primary flotation step to such an elaborate extent. Not only was there a decrease in the amount of rubber hydro-
TABLEVI, SOXE QUALITYTESTSOF CRUDERUBBERFROM FRESH SHRUB A S D FROhI SHRCB S T O R E D F O R V A R Y I X G PERIODS O F TIME Insolu-
Shrub Treatment Fresh Shrub field-cured for 1 week baled, then stored for weekly periods of
bles in Crude Rubber
6.1 6 6
7.0 7.6 7.9 7.8 8.1 8.3
7.9
Parts Resin
~ e 100 r
-Parts Hydrocarbon 28 9 26 7
23.7 25.7 26.0 25.8 25.1 25 7 23.8 -0.59 -0 30
n4olecular Weight ( X 1000)
Mooney ML, 21Z0,F., 5 Ahn.
150 104 112 111 93 104 101 97
53.0 45.0 47.0 46.0 11.0 45.0 43.0 43.0 37.5 89 -0.8Oa -0.82“ -4.8ga -1.24“
Correlation coefficient, r +O 92‘ +O 23‘ Regression coefficlent, 6 a Value greater than required for significance a t 1 % level of probability.
ACKNOWLEDGMENT
The authors are grateful to Ruth 5’. Crook, R. H. Taylor, William P. Ball , and James W.Meeks of this laboratory for their assistance in chemical analyses, 8Iooney viscosity, and molecular weight determinations. They are also indebted to Eleanor C. Taylor, of the Bureau of Plant Industry, Soils, and A4gricultural Engineering, for her work in statistical analyses. LITERATURE CITED
(1) Addicott, F. T., A m . J . Botany, 32, 250-6 (1945). (2) Allen, P. J., and Emerson, Ralph, IKD.EKG.CHEM.,41, 346
(1949). (3) Cumming, J. M., and Chubb, R. L , Chem. Ce- Met. Eny., 53, 125-6 (1946). (4) Davis, C. C., and Blake, J. T., “The Chemistry and Technology of Rubber,” BP. 226-31, New York, Reinhold Publishing Corp., 1947. (6) Eighty-Second Congress, 1st Session, Twenty-Ninth Rept. Preparedness Subcommittee, Committee on Armed Services, U. 8. Senate, Third R e p t . on Rubber, 2 (1951). (6) Forest Service, U. S. Dept. Agr., India Rubber World, 109, 363-6, 370 (1944). (7) Lloyd, F. E., Carnegie Inst. W a s h . Pub.,No 139 (1911). (8) Sishimura, h/l. S., Hirosawa, F. N., and Emerson, Robert, IND. ENG.CHEM.,39, 1477 (1947). (9) Snedecor, G. W., “Statistical Methods,” 4th ed., Ames, Iowa, Iowa State College Press, 1946. (10) Spence, D., and Caldwell, X I . L., IND, ENG CHEhz., ANAL. ED.,5 , 371 (1933). (11) Taylor, K. W., Econ. Botany, 5 , 265-73 (1951). (12) Taylor, K. W., and Chubb, R. L., IND.ENQ.CHEM.,44, 883 (1952). (13) Taylor, R . H., Fielding, J. H., and Mooney, M., ASTM Symposium on Rubber Testing, 1947. (14) Tysdal, H. M., Agron. J., 4 2 , 351-5 (1950). RECEIVED for review March 12, 1951.
BCCEPTED October 21, 1951.