Effects of firing temperature on the fate of naturally occurring organic matter in clays

Effects of firing temperature on the fate of naturally occurring organic matter in clays

Journal of Archaeological Science 1988,15,403%414 Effects of Firing Temperature on the Fate of Naturally Occurring Organic Matter in Clays Jessica S...

923KB Sizes 0 Downloads 56 Views

Journal of Archaeological

Science 1988,15,403%414

Effects of Firing Temperature on the Fate of Naturally Occurring Organic Matter in Clays Jessica S. Johnson’“, Jeff Clarkb, Sari Miller-Antoniob, Michael B. Schifferd and James M. Skibod (Received

29 June 1987, revisedmanuscript

accepted21

December

Don Robins’,

1987)

Experiments were carried out to test the effects of firing temperature on the fate of naturally occurring organic matter in clay. These results have important applications for the radiocarbon dating of archaeological ceramics. De Atley (1980) originally discussedthe effects of carbon derived from coal or other old organic material in clays. Our experiments yielded quantitative measurements of carbon remaining in the ceramic after firing. Two commercial clay compositions were examined; low organic and high organic. Elemental analysis shows that carbon remained in ceramics fired at temperatures of SO&lOOO”C,well exceeding the maximum firing temperatures of most primitive techniques. Organic burn-out is also dependent on the initial percentage of organics in the clay. Similar results for firings in oxidizing and non-oxidizing atmospheres suggest that organic burn-out is not only a combustive process but also a pyrolytic one. Keywords: CERAMICS, RADIOCARBON

EXPERIMENTAL ARCHAEOLOGY, SPECTROSCOPY.

DATING, ORGANICS, ELEMENTAL ANALYSIS, INFRA-RED

Introduction Direct dating of organic material in the matrix of archaeological ceramics has been attempted using standard radiocarbon techniques (e.g. Evans & Meggers, 1962; Stuckenrath, 1963; Taylor & Berger, 1968; Gillespie et al., 1984). However, the quantity of pottery that had to be processed to yield sufficient carbon for conventional radiocarbon dating was often prohibitive. With the advent of accelerator (AMS) radiocarbon dating, capable of assaying very small amounts of carbon, lack of material is no longer an impediment. As a result, the direct dating of archaeological ceramics is apt to increase in the future (e.g. Johnson et al., 1986). Before this practice becomes commonplace, a greater “Western Archeological and Conservation Center, National Park Tucson, Arizona 85717, U.S.A. ‘Department of Anthropology, The University of Arizona, Tucson, 85721, U.S.A. ‘18 Chequers Square, Uxbridge, Middlesex UB8 lNE, England. dLaboratory of Traditional Technology, Department of Anthropology, University of Arizona, Tucson, Arizona 85721, U.S.A. *Present address: University College of London, Institute of Archaeology, Gordon Square, London WClH OPY, England.

Service, Arizona

The 31-34

403 03054403/88/040403

+ 12 $03.00/O

0 1988 Academic

Press Limited

404

J. S. JOHNSON ET AL.

understanding is needed for the diverse sources of organic carbon in clay and its response during firing of the ceramic, for these can clearly influence the results of radiocarbon dating. De Atley (1980) reviewed previous radiocarbon dating of ceramics, identifying a disturbing pattern of anomalously early dates. These initial attempts assumed that all carbon in the clay matrix was culturally derived. As a possible mechanism to account for the older-than-expected dates, de Atley posited the presence of naturally derived old carbon. She pointed out that old carbon of geological origin in sedimentary clays would “dilute” the culturally introduced carbon and yield excessively early dates. Although this mechanism is quite plausible, we do not know the conditions under which geologically introduced carbon can survive firing. Clearly, a better understanding of carbon loss during firing is essential to provide a foundation for radiocarbon dating of archaeological ceramics. In the present study, we quantify the survival of naturally occurring carbon that has been subjected to a range of firing temperatures in both oxidizing and non-oxidizing atmospheres. The literatures of archaeology, ceramic engineering and ceramic art contain a variety of statements about the temperature at which carbon loss takes place during firing (see Table l), as well as some of the factors that might influence the process. However, few of the works provide relevant experimental data. In most studies where such data are presented, the experimental conditions and measuring techniques are appropriate for engineering ceramics but not for those customarily encountered by the archaeologist. This project was designed to simulate primitive firing techniques in a controlled environment. “Primitive” firing is defined as a heating process for ceramics that rises quickly to a relatively low maximum temperature (Shepard, 1980: 424) which is maintained for a short period of time. For example, Colton’s (1951) study of Hopi firing techniques indicated that 35-75 min was required to reach the maximum temperature of 885°C. Our firings were carried out in a furnace to control external variables such as firing temperature, atmosphere and soak time, all of which might affect the amount of carbon in the fired ceramic. In these firings, the maximum temperature was quickly achieved and subsequently held for 30 min. Although we have attempted to simulate “primitive” firing techniques, the laboratory simulation may differ in one important respect. The 30 min soak used in the experiment is longer than the duration of maximum temperature found in many open firings, probably resulting in a more thorough loss of carbon in the experimental briquets. Retention of original carbon in clay is apt to be more of a problem under real primitive firing conditions. Therefore, our findings on carbon retention can be regarded as very conservative. Organic Material in Clay It is first necessary to appreciate the five general sources of organic matter in archaeologically recovered pottery (Gabasio et al., 1986: 712-13). First, organic matter is naturally present in the clay itself. All natural clays contain organic matter, with percentages of less than 1% for some residual clays and up to 17% for some secondary clays (Worrall, 1968: 33). Second, organic matter can be added to clay as temper, such as the fibres added by Late Archaic potters in the eastern United States (Reid, 1984). Third, carbon can be introduced during pottery manufacture by (1) deposition of material from the fuel during firing (Gabasio et al., 1986) and (2) smudging and applying of resins or some other organic matter (e.g. Gayton, 1929; Longacre, 1981). Fourth, organic matter can be deposited during use as soot on cooking pots (e.g. Hally, 1983) or absorbed from vessel contents. Fifth, in principle, humic acids and other organic matter can be absorbed by ceramics from the depositional environment (cf. Schiffer, 1987: 159961). Previous researchers, using radiocarbon techniques to date ceramics, have assumed that all carbon in the paste

EFFECT

OF FIRING

TEMPERATURE

Table I. Archaeological, ceramic engineering discuss loss of carbon from clay during$ring Temperature

Source Binns

(1967:

116)

Bole&Jackson

173)

Bose (1942) in Cardew Denio (1980: 273)

(1969:

Duma

& Lengyel(l970:

Goffer

(1974:

Green Halley

& Stewart (1953: (1983: 11)

Hamer

(1975:

Kingery

Kraner

75)

200-500°C

Ryan

& Fritz

(1978: (1978:

Rye (1981:

233)

35&65O”C 500°C

211)

et al. (1976:

70&l 503)

(1929:

237) 37-8)

4o(t8oo”c 200-1ooo”c 200-900°C

118)

McNamara & Comeford Morgan (1936: 10) Nelson

460°C 210)

150°C

15&4OO”C

13) (1945:

30)

Completed at 60~800”c < 500°C - 1850°F 510-705°C 300-700

+ “C

20&9OO”C

108)

and ceramic

art literature

sources

that

Comments

Red hot

(1924:

405

Shepard

(1936:

398)

350-950°C

Shepard

(1980:

21)

225-800°C

“Some native clays are rich in carbonaceous matter and therefore require a longer time for complete oxidation” “carbon is burnt out completely at 460°C” and “it is burnt out more rapidly the more rapid the rate of heating” “the ignition temperature of carbon” “assuming an adequate supply of air is present” and “depending on the types of organic residues present” “pottery materials containing organic substances gradually lose their carbon content if fired under oxidizing conditions at a relatively low temperature” “If heated under reducing conditions (when oxygen is not available), organic compounds are not oxidized but are charred (i.e, partly burned), leaving a residue of elemental carbon” “carbonaceous matter burns out up to 650°C” “When heated in an oxidizing atmosphere, the organic material in such vessels will decompose at temperatures above approximately 200°C. The carbon released by this process will in turn oxidize and dissipate as carbon dioxide when temperatures reach approximately 500°C” “The so-called burning out periods take place between 700°C and 1150°C” “under normal conditions organic materials char at temperatures above 150°C and burn out at temperatures ranging from 300-4OO”C” “oxidation is dependent on temperature (not time) and porosity of the body” “burnout of binders (temper)” “by contact with reducible materials such as Fe,O,, MnO,, SO,” “During this ‘water smoking’ period gases from organic material leave the body” “the temperature and ease of removal depending on the type of organic material present and the rate of heating” “Under suitable oxidizing conditions, all carbon will be removed by about 900°C except graphite, which can resist oxidation up to about 1200°C” “The rapidity with which carbon is removed during firing depends upon its condition and upon the density of the clay” “the most effective temperatures are. . generally between 700-800°C”

was of the second, third and fourth types; that is, it had been introduced culturally when the vessel was made or used. Our experiment examines only naturally occurring organic material and its loss during firing. This organic matter can be introduced from several sources, as Shepard (1980: 19) notes: Vegetable matter is common in surface clays of recent origin; bituminous and asphaltic matter occurs in shales and clays associated with coal measures; and lignite is common in Cretaceous clays. Anthracite also occurs, and some clays and shales contain organic matter that can be extracted with water or other solvents.

406

J. S. JOHNSON

ET AL.

Ceramic engineers have concluded that one source of naturally occurring organic material in clays is coal or a closely related substance. Ball clays, sedimentary clays composed primarily of kaolinite (Rice, 1987: 51), often occur in association with beds of lignite, an immature brown coal intermediate between peat and bituminous coal. Lignite has two principal constituents: (1) waxes and resins (long-chained hydrocarbons), and (2) a partly oxidized aromatic substance from the lignin component of wood (termed humic acid). During deposition, finely divided particles of the lignite become adsorbed on the clay particles. In addition, lignite contamination can occur during mining of the clay (Worrall, 1956: 689). Ball clays have been found with up to 17% organic material but the majority have 1% or less (Worrall, 1968: 33). Fire clay, a refractory type clay, is found alternately bedded with bituminous coal (Worrall, 1968: 35). Bituminous coals are older deposits in which the humic acids have undergone further condensation to graphitic condensed ring compounds with the waxes and resins loosely combined. These types of coal may also be adsorbed on the fire clay molecules as well as contaminating the bulk. Additionally, clay contamination might be caused by particulate coal or soluble organic compounds transported by groundwater percolation through overlying layers of coal (Tankersley et al., 1987: 326). Soil humus can also be incorporated in clay beds. However, this source of carbon has probably a lesser effect on radiocarbon dates of ceramics than older, less volatile coal-related sources. Sample Preparation and Analysis In order to study the rate of carbon loss in clay during heating, we fired clay briquets under controlled conditions in a furnace. Then, using standard elemental analysis techniques, samples of the briquets were measured for carbon and hydrogen content. In the unfired clays, percentage of carbon as carbonate was also measured. In addition, infra-red spectroscopy was used to examine the chemical constituents of the organic matter and their changes during firing. Sample preparation involved the fabrication of two sets of ceramic briquets from two Westwood Ceramics Supply Co. clays (see Appendix 1). The percentage of organics in each clay was experimentally determined during this analysis. A dry weight of 900 g of Kentucky Special, a ball clay that has about 5% organics, was mixed with 450 ml of distilled water and rolled out on paper into a 9.5 mm thick slab. (Paper fibres adhering to the surface of the clay briquets would not affect our conclusions, since paper burns at or below approximately 25O’C.) While still wet, this slab was cut into 3 cm square briquets and transferred to a plaster drying bat. The second set of briquets was made from pre-moistened Westwood Porcelain clay consisting of l-2% organic material. Using the same technique as above, 1200 g of the pre-moistened porcelain was made into briquets. Both low and high organic briquets were fired together in an oxidizing environment using a Neycraft JFF 2000 Manual Fiber Furnace. A second set of low and high organic briquets was fired in a non-oxidizing environment using a Superamics gas-fired raku kiln (Troncone, n.d.). A non-oxidizing atmosphere was assumed when the gas flame had an orange colour and the smell of unburned fuel was present. However, there was no quantitative measurement of the oxygen present in the kiln during firing. In both environments, the briquets were fired at 200°C intervals from 200-1000°C. They were taken up to the desired temperature, from ambient, and held there for 30 min. It should be noted that the heat-up times varied by firing atmosphere. For example, the oxidizing kiln took 18 min to reach 200°C and 50 min to reach 1000°C. The non-oxidizing furnacereached a maximum temperature more quickly; 1 min to reach 200°C and 23 min to reach 1000°C. The briquets

EFFECT OF FIRING TEMPERATURE Table 2. Results and non-oxidizing

of elemental analysis atmospheres

407

on high and low organic

Sample High organic/unfired Low organic/unfired

clays$red

%C

%H

3.41 0.16

1.32 0.65

200°C firing High organic/oxidizing High organic/non-oxidizing Low organic/oxidizing Low organic/non-oxidizing

atmosphere atmosphere atmosphere atmosphere

3.30 3.36 0.14 0.15

1.21 1.26 0.71 0.68

400°C firing High organic/oxidizing High organic/non-oxidizing Low organic/oxidizing Low organic/non-oxidizing

atmosphere atmosphere atmosphere atmosphere

1.43 1.47 0.03 0.04

0.93 0.94 0.59 0.67

600°C firing High organic/oxidizing High organic/non-oxidizing Low organic/oxidizing Low organic/non-oxidizing

atmosphere atmosphere atmosphere atmosphere

0.69 0.94 0.009 0.014

0.20 0.14 0.11 0.07

800°C firing High organic/oxidizing High organic/non-oxidizing Low organic/oxidizing Low organic/non-oxidizing

atmosphere atmosphere atmosphere atmosphere

0.25 0.28 0.035 0.010

0.09 0.07 0.033 0.020

1000°C firing High organic/oxidizing High organic/non-oxidizing Low organic/oxidizing Low organic/non-oxidizing

atmosphere atmosphere atmosphere atmosphere

0.009 0.026 0.010 0.004

0.042 0,037 0,036 0,028

in oxidizing

%C ascarbonate

0.006 0.002

fired in the oxidizing atmosphere were allowed to cool to room temperature before being removed from the kiln; those fired in the non-oxidizing atmosphere were allowed to cool to 200°C before removal. The primary analytical step consisted of an elemental analysis of the briquets for carbon and hydrogen content. Each briquet was finely ground by mortar and pestle, and a 1 g sample was analysed in an enclosed combustion chamber. The results are presented in Table 2 and Figures 1 and 2. To ensure that carbonates were not a major source of carbon in the unfired clay, percentage of carbon as carbonate was also measured by elemental analysis. The high organic clay contained 0.006% carbon as carbonate; the low organic clay contained 0.002%. These small percentages of carbonate do not affect the measurement of total carbon in the clays. In addition, we obtained infra-red spectra on two high organic samples. A sample of unfired high organic clay and a sample fired to 400°C in an oxidizing environment were extracted in ether and acetone solvent systems. The solvents were subsequently removed under vacuum. A sample smear was prepared and spectra were obtained using a Perkin-Elmer 599B Infrared Spectrophotometer.

408

ET AL.

J. S. JOHNSON 3.6

High organic otmosphere

2 .$

2.2

3 2

2.0

s 9 z 0 ae

non-oxidizing

I.8 I. 6 1.4 I. 2 I.0 0.8 0.6 0.4

0

200

400 Firing

600 temperoture

Figure 1. The effect of firing temperature determined by elemental analysis.

800

1000

(“Cl

and kiln atmosphere

on the loss of carbon

Results The elemental analysis indicates that carbon loss in the high organic clay has begun by 200°C and continues to 1000°C. By comparison, carbon in the low organic clay briquets is essentially gone by 600°C. The most dramatic loss for both clays occurs from 2OWlOO”C (Figure 1). The graph of hydrogen present in both clays suggests that the greatest loss of this element, bound chemically either to carbon or inorganic constituents, is in the 400-600°C range and is essentially complete by 800°C (Figure 2). Finally, there appears to be little variation in the amount of carbon and hydrogen loss between oxidizing and non-oxidizing environments. This unexpected result is discussed below. Infra-red spectroscopy on two of our samples provided additional information on the nature of the chemical bonds present in the organic constituents of the high organic clay. In the spectrum for the unfired clay, a greater variety of chemical bonds was present. The unfired clay spectrum showed strong peaks for the C - H (2900 cm- ‘), C - 0 (1110 cm- ‘) and C = 0 (1700 cm- ‘) bonds, representing oxygenated organic compounds. By contrast, the spectrum of soluble organic material extracted from the high organic clay fired at 400°C did not contain peaks in the C - 0 and C = 0 regions; only strong absorption in the C-H band is apparent. Thus, more reactive oxygenated compounds have thermally degraded by 400°C (diffusing from the clay matrix in the form of carbon dioxide), leaving

EFFECT

OF FIRING

TEMPERATURE

igh organic

409

non-oxidizing

I.0 z .IF $ 2

0.9 0.8

5

0.7

z

0.6

I” s

0.5

oxidizing

atmosphere

0.4

0

200

400 Firing

600

temperature

Figure 2. The effect of firing temperature gen determined by elemental analysis.

800 (“C)

1000

and kiln atmosphere

on the loss of hydro-

behind the less reactive long-chain hydrocarbons and the non-volatile byproducts of thermal degradation. The infra-red spectroscopy (IR) and elemental analysis suggest the following thermal regime for the loss of organic matter in the high organic clay. (1) I-200°C. Elemental analysis shows that little organic material has been removed in this temperature range (Figure 1). (2) 200400°C. Infra-red spectroscopy of extractable organic material demonstrates the complete loss of oxygenated organic compounds. (3) 400-600°C. Elemental analysis indicates that the greatest loss of hydrogen occurred in this range (Figure 2). This may represent the removal of less reactive hydrocarbons and the byproducts of initial thermal degradation, which would exhibit greater thermal stability than the organics containing oxygen. (4) 600-800°C. Elemental analysis shows that essentially no organically bound hydrogen is present after 800°C (Figure 2). Thus, all hydrocarbons have been removed or thermally degraded into more condensed compounds. (5) SO&1000°C. Since essentially all hydrogen has been lost, elemental analysis suggests that the carbon remaining after 800°C is probably in the form highly condensed aromatic compounds structurally resembling coal, which could have been present originally in the unfired clay or produced during firing as non-volatile byproducts of thermal degradation. Discussion In the present study, the loss of naturally occurring organic material in clays during firing was shown to be dependent on two variables.

410

J. S. JOHNSON

ET AL.

Clay composition The temperature range in which organic removal occurred was directly related to the amount of organics present in the clay (see also Gabasio et al., 1986). Carbon loss in the high organic clay occurred over a wider temperature range and was complete at a higher temperature. Clay porosity, inorganic composition and the amount of chemically bonded oxygen present may also affect the rate of organic removal (Troncone, n.d.). Firing temperature The maximum firing temperature of the clay was an essential factor in the degree of organic removal. This study indicated that, while the majority of the organic material was removed between 200 and 800°C in the high organic clay, some remained until the firing temperature reached the 900-1000°C range. Complete loss of organics in the low organic clay occurred at a lower temperature (c. 600°C). Firing time is also an important variable in the degree of organic removal. After surveying the ethnographic literature, we chose 30 min to bracket the range of soak times used in primitive open firings. Actual soak times would obviously vary from society to society and between potters. Nicholson & Ross (1970) report that the importance of firing time on the extent of organic combustion in clays is probably insignificant after 3040 min at temperatures above 500°C. Clearly, this variable does play a significant role in the combusion of organics in the clay matrix at lower temperatures or for soak times of less than 30 min. It is interesting to note that in the present experiment, kiln atmosphere had little effect on the extent of organic removal. Briquets fired in the non-oxidizing atmosphere retained only slightly more organics at the various firing temperatures (Figure 1). This nonoxidizing atmosphere is not the same as a “reduction” firing or “smudging” used in primitive technologies, where carbon from incompletely burned fuel is deposited on the ceramic. In this experiment, it was important to control the sources of carbon in the fired ceramic. Therefore, firing in a non-oxidizing environment simply allowed us to compare the effects of atmospheric oxygen on the loss of carbon in the fired ceramic. Both nonoxidizing conditions and shorter heat-up time in the raku furnace were expected to promote less organic loss, yet that did not occur. Several possible explanations of this counter-intuitive result follow: (1) There was sufficient oxygen in the non-oxidizing furnace to combust the organic matter in the clay to the same extent as in the oxidation furnace. (2) The clay itself contained enough organically and inorganically bound oxygen to combust the carbonaceous material to the same extent in both oxidation and nonoxidation atmospheres in the given firing time. (3) Enough oxygen was present in the kiln to combust the organic material in the clay as the briquets were allowed to cool to room temperature. (4) Pyrolysis, the heat-induced decomposition of non-volatile organic compounds in the absence of oxygen, played a large role in the removal of organics from the clay matrix. The end products of this reaction are volatile short-chained hydrocarbons (e.g. methane, ethane, propane) and condensed non-volatile aromatics. The latter would remain in the clay matrix until further pyrolysis (requiring higher temperatures) took place and may be responsible for the “carbon core” so common in primitive ceramics. Thus, organic pyrolysis, in addition to combustion, may contribute to the removal of organic matter from ceramics, especially in later stages of firing. In general, the temperature ranges for the removal of organics in the literature (Table 1) are either in agreement with our results or lower, but only in one case higher. This range is clay source-dependent, however, and the high organic clay used in this study probably represents an upper limit. Kraner & Fritz (1929) examined a series of organic combustion

EFFECT

OF FIRING

TEMPERATURE

411

regimes for various clays based on carbon dioxide evolution. They found that although the individual rates for each clay are unique, the general temperature range for the maximum combustion of organics is 200-8OO”C, with complete combustion occurring just over 800°C. Note that this last figure is approximately 100°C lower than our results on high organic clay. Nicholson & Ross (1970) examined the variability of organic combustion in clay with firing time and temperature. Their findings also support our experimental results. They found that the majority of carbonaceous material cornbusted between 300 and 7OO”C, and complete combustion occurred just above 900°C. It is important to note that earlier experiments measured carbon dioxide evolution during firing and not the organic material remaining in the fired clay. Thus, they assume that all mechanisms for organic removal from clay are combustive processes in which carbon dioxide is evolved. The validity of this assumption is questionable in view of our preliminary infra-red spectroscopy results, which show that pyrolysis in the absence of oxygen may also play a significant role in organic burn-out. As noted above, the gaseous products of this process would be short-chained hydrocarbons rather than carbon dioxide. The direct measurement of residual carbon content in the present study eliminates this potential source of error. Conclusions The results obtained in this experiment indicate that care. must be exercised in the radiocarbon dating of ceramic material. This study demonstrates that organics naturally occurring in clays can survive firing temperatures above 9OO”C, which is beyond the temperatures used by many prehistoric ceramic technologies. If sedimentary clays containing older carbonaceous material such as lignite, bituminous coal and anthracite were used in pottery making, then older-than-expected dates would be obtained (de Atley, 1980: 989). Unfortunately, the present study indicates that these older, more chemically inert organics are the last to burn out of the clays upon firing (see also Gabasio et al., 1986). Gabasio et al. (1986: 717) propose that controlled combustion could be used to separate naturally occurring carbon from that added during manufacture. Although this possible remedy merits careful study, the authors’ claim that “carbon acquired during firing may outweight all other carbon sources in potsherds” (Gabasio et al., 1986: 717) is probably premature in view of the limited experiments carried out so far on this issue. De Atley (1980) suggests that one can correct for the “old carbon” problem by subtracting the radiocarbon date of the source clay, if it is known, from that of the fired ceramic. Regrettably, such a correction procedure does not solve the problem because some of this “old carbon” inevitably would have been removed during firing. If the geological source of the clay is known, contamination by “old carbon” can be identified. Ideally, the firing temperature and length of firing time of the ceramics should be known for calibrating the original radiocarbon dates by dating replicative firings of clay source samples. Recent work in electron spin resonance (e.g. Robins et al., 1983) may be useful in determining these variables. Restricting the carbon sample for dating to demonstrable organic temper particles is one obvious remedy to avoid contamination by “old carbon”. Direct radiocarbon dating of ceramic sherds can be a valuable tool for the archaeologist because it eliminates the problem of finding preserved organic remains in association with ceramics. However, as this study indicates the limitations of such direct dating must be identified and confronted before it can become a reliable analytical technique, Further investigation of the following issues is warranted. The possibility that organic burn-out is independent of firing atmosphere was suggested by our results, thus indicating a possible role played by organic pyrolysis. This counter-intuitive finding requires

412

J. S. JOHNSON ET AL.

further inquiry using a kiln in which the amount of oxygen present can be controlled more precisely. Characterization of the organics remaining in the clay at various firing temperatures, perhaps using mass spectroscopy, would be helpful for clarifying the contributions of combustive and pyrolytic processes during the burn-out of carbonaceous material. Finally, further studies that quantify the effects of soak time on carbon loss would complement this experiment. Acknowledgements This research was supported by a grant to M. B. Schiffer from the National Science Foundation (BNS-84-19935) and by gifts from Harcourt Brace Jovanovich, the Stephen Tyler Fund (of the University of Arizona Foundation) and Watson Smith to the Laboratory of Traditional Technology. We are grateful to Labline Inc. and Setra Systems Inc. for donation of equipment used in the experiment. Also, we thank E. Sherrill, President of Westwood Ceramics Supply Co., for supplying information on the composition of the clays. Desert Analytics, Tucson, Arizona, conducted the elemental analysis of our clay samples. Special thanks go to Brigid Sullivan for the use of space and equipment in the Conservation Laboratory of the Western Archeological and Conservation Center. We are grateful to Chris Carr, Alfred E. Johnson, Ben Nelson and John Speth for their suggestions and comments on the paper. Finally, we thank Lisa Young for her technical assistance and support during the preparation of the manuscript. Senior authorship of this paper is shared by Jessica S. Johnson, Jeff Clark and Sari Miller-Antonio. References Binns, C. F. (1967). The Potter’s Craft. New York: Van Nostrand Company. Bole, G. A. & Jackson, F. G. (1924). The oxidation of ceramic wares during firing. I. Some reactions of a well known fire clay. Journal of the American Ceramic Society 7, 163-174. Cardew, M. (1969). Pioneer Pottery. New York: St Martin’s Press. Colton, H. S. (1951). Hopi pottery firing temperatures. Plateu, Museum of Northern Arizona 24, 73-76.

de Atley, S. P. (1980). Radiocarbon dating of ceramic material: progress and prospects. Radiocarbon

22,98&993.

Denio, A. A. (1980). Chemistry for potters. Journal of Chemical Education 57(4), 272-275. Duma, G. & Lengyel, I. (1970). Mesocstat pot containing red blood pigment (haemoglobin). Acta Archaeologica

Academic Scientiarium

Hungaricae

22,69-93.

Evans, C. & Meggers, B. (ly62). Use of organic temper for C-14 dating in lowland South America. American

Antiquity

28,243-245.

Gabasio, M., Evin, J., Arnal, G. B. & Andrieux, P. (1986). Origins of carbon in potsherds. Radiocarbon 28(2A), 711-718. Gayton, A. H. (1929). Yokuts and Western Mono pottery-making. In American Archaeology and Ethnography, Vol. 24. Berkeley: University of California, pp. 239-255. Gillespie, R., Gowlett, J. A. J., Hall, E. T. & Hedges, R. E. M. (1984). Radiocarbon measurement by accelerator mass spectrometry. Archaeometry 26, 18-20. Goffer, Z. (1974). Archaeological Chemistry (Advances in Chemistry). Washington: American Chemical Society. Green, A. T. & Stewart, G. H. (1953). Ceramics: a Symposium. Stoke-on-Trent: The British Ceramic Society. Hally, D. J. (1983). Use alteration of pottery vesselsurfaces: an important source of evidence for the identification of vesselfunction. North American Archaeologist 4,3-26. Hamer, F. (1975). The Potter’s Dictionary of Materials and Technique& London: Pitman Publishing.

EFFECT

OF FIRING

TEMPERATURE

413

Johnson, R. A., Stipp, J. J. & Tamers, M. A. (1986). Archaeologic sherd dating: comparison of thermoluminescent dates with radiocarbon dates by beta counting and accelerator techniques. Radiocarbon 28(2A), 719-725. Kingery, W. D., Bowen, H. K. & Uhlmann, D. R. (1976). Introduction to Ceramics. New York: John Wiley & Sons. Kraner, H. M. & Fritz, E. H. (1929). The rate of oxidation of porcelain and ball clays. Journal of the American Ceramic Society 12(l), 1-13. Longacre, W. A. (1981). Kalinga pottery: an ethnoarchaeological study. In (I. Hodder, G. Isaac & N. Hammond, Eds) Patterns ofthe Past: Studies in Honor of David Clarke. Cambridge: Cambridge University Press, pp. 49-66. McNamara, E. P. & Comeford, J. E. (1945). Classification of natural organic binders. American Ceramic Society Bulletin 28,25-3 1. Morgan, W. M. (1936). Oxidation and loss of weight of clay bodies during firing. University of Illinois Engineering Experimental Bulletin 284,740. Nelson, G. C. (1978). Ceramics: a Potter’s Handbook, 4th edition. New York: Holt, Rinehart & Winston. Nicholson, P. S. & Ross, W. A. (1970). Kinetics of oxidation of natural organic material in clays. Journal of The American Ceramic Society,53, 154-158. Reid, K. C. (1984). Nebo Hill: Late Archaic prehistory on the Southern Prairie Peninsula. Publication in Anthropology 15. Lawrence: University of Kansas. Rice, P. (1987). Pottery Analysis: A Sourcebook. Chicago: University of Chicago Press. Robins, V. D., de1 Re, C., Seeley, N. J., Davis, A. G. & Hawari, J. A.-A. (1983). A spectroscopic study of the Nimrud Ivories. Journal of Archaeological Science 10,385-395. Ryan, W. (1978). Properties of Ceramic Raw Materials. Oxford: Pergamon Press. Rye, 0. S. (1981). Pottery Technology: Principles and Reconstruction. Manuals on Archaeology 4. Washington: Taraxacum. Schiffer, M. B. (1987). Formation Processes of the Archaeological Record. Albuquerque: University of New Mexico Press. Sharratt, E. & Francis, M. (1943). The organic matter of ball clays. Part I. Transactions of the British Ceramic Society 42, 111-121. Shepard, A. 0. (1936). The technology of Pecos pottery. In The Pottery of Pecos, Vol. 2. New Haven: Yale University Press. Shepard, A. 0. (1980). Ceramics for the Archaeologist. Washington: Carnegie Institution of Washington, Publication 609. Stuckenrath, R. (1963). University of Pennsylvania radiocarbon dates VI. Radiocarbon 582-103. Tankersley, K. B., Munson, C. A. & Smith, D. (1987). Recognition of bituminous coal contaminants in radiocarbon samples. American Antiquity 52,3 18-330. Taylor, R. E. & Berger, R. (1968). Radiocarbon dating of the organic portion of ceramic and wattle-and-daub house construction material of low carbon content. American Antiquity 33, 363-366. Troncone, S. M. (n.d.). An experiment in the production of carbon coring in lowfire ceramics. Manuscript in preparation. On file at University of Arizona, Laboratory of Traditional Technology, Department of Anthropology. Worrall, W. E. (1956). The organic matter in clays. Transactions of the British Ceramic Society 55, 689-705. Worrall, W. E. (1968). Clays: Their Nature, Origin and General Properties. London: MacLaren & Sons.

414

J. S. JOHNSON

ET AL.

Appendix 1. Basic composition supplied by Westwood Ceramic

datafor Supply

the clays used in the experiment Company)

Unknown

Low organic clay (WCS38 Porcelain) Silicon dioxide Aluminium oxide Iron oxide Manganese oxide Titanium dioxide Calcium oxide Magnesium oxide Sodium oxide Potassium oxide Sulphur trioxide Phosphorous pentoxide High organic clay (Kentucky Special) Silicon dioxide Aluminium oxide Iron oxide Manganese oxide Titanium dioxide Calcium oxide Magnesium oxide Sodium oxide Potassium oxide Sulphur trioxide Phosphorous pentoxide

mixture

l/2 Champion and l/2 Challenger

EPK-Edgar Plastic Kaolin

53.96% 29.34 0.9e 0.02 1.64 0.37 0.30 0.12 0.28 0.03 0.15

45.91% 38.71 0.42 0.12 0.34 0.09 0.12 0.04 0.22 -

50.5% 28.7 0.9 1.5 0.4 0.4 0.2 0.9

-

(information

of G-200 Feldspar

67.40% 18.50 0.09

1.15 3.24 9.40 -

200 Mesh Silica

99.8% 0,050 0.02 0.020 0.008 0.007 0.008 -