“Our Ancestors Were Material Scientists”

Abstract

The quest for a comprehensive understanding of Africa’s indigenous technology has been an important intellectual agenda in Black Studies. In some instances this interest has tended to be speculative and derivative because most Black Studies scholars are not trained in the disciplinary fields that are relevant for investigating Africa’s indigenous technology through primary field and laboratory research. Collaboration between Black Studies scholars and those in the physical sciences is therefore important in order to develop new insights into Africa’s indigenous technology. One of such collaborations has led to a new archaeological and geochemical evidence for primary glass manufacture in Yorubaland. Based on the recent archaeological finds of glass artifacts from Osun Grove (Osogbo, Nigeria), we present the laboratory data that lead us to the conclusion that the Yoruba of West Africa developed a unique glassmaking technology that lasted till the seventeenth century.

Keywords

African indigenous technology Yoruba glass glass beads geochemical analysis Osun Grove Nigeria

Introduction

Glass beads have special cultural significance for the Yoruba, whose ancestral homeland includes the present-day southern Togo, southern Benin, and southwestern Nigeria, in West Africa (Figure 1). For most of Yoruba history, especially from the 11th through 18th centuries, glass beads were the ultimate objects of wealth, primary insignia of political status, badges of royalty, and emblems of spiritual authority (Ogundiran, 2002). Archaeological evidence demonstrates the existence of a vast glass bead industry in the Yoruba ancient city of Ile-Ife from about 1000 to 1500 ad (Eluyemi, 1987; Willett, 1967, 1977). Many scholars had thought that this industry was a secondary one, using recycled and reheated glass obtained from medieval Europe and the Islamic world (e.g., Willett, 1977). There was, however, a small group of dissenting voices who speculated on the possibility of indigenous primary glass manufacture in Ile-Ife (Eluyemi, 1987). A leading figure who represented this view was the Late Dr. Omotoso Eluyemi who declared on many occasions during his lectures that “our ancestors were material scientists.”¹ An archaeologist and a high-ranking chief of Ile-Ife, Eluyemi was referring to the complex process of glass production in the ancient city. This oft-repeated statement was meant as a counterpoint to the then-popular view that the glass bead makers in Ile-Ife were importing glass from Europe and Islamic world and re-melting the glass to make beads between 11th and 15th centuries. But until recently, the geochemical evidence needed to exorcise this ghost of Eurocentric thought about Africa’s indigenous technological skills was not available. Hence, his claim of indigenous primary glass manufacture among the Yoruba was not taken seriously both in the sciences and in the historiography.

Figure 1.

Ancestral Yorubaland, West Africa.

However, over the past decade, the compositional analysis of glass-lined crucibles, glass cullet (unshaped glass objects), and glass beads of different colors, retrieved from excavation sites and groves in Ile-Ife, have given tentative chemical evidence for primary glass manufacture in the ancient city (e.g., Ige, 2009, 2010; Lankton, Ige, & Rehren, 2006). J. W. Lankton, O. A. Ige, and Th. Rehren used scanning electron microscopy with Energy Dispersive X-ray Spectroscopy (SEM-EDS), electron-probe microanalysis (EPMA), and X-ray fluorescence (XRF) to investigate the elemental composition of glass-working crucibles, cullet, and drawn glass beads (totaling 17 samples) from the Olokun Grove and Ita Yemoo sites, as well as from museums and private collections in Ile-Ife. The majority of these Ife materials show unusually high-lime, high-alumina (HLHA) content, greater than the proportions of alumina and lime found in ancient Islamic, European, and Asian centers of glass production. Comparing their data with similar analyses of glass beads in other parts of West Africa, Lankton et al. (2006: 136) tentatively concluded that the HLHA beads are “evidence for a glassmaking tradition unique to West Africa, and possibly unique to southern Nigeria and to the Yoruba culture”. The authors also point to the local availability of raw materials that would have been used to produce the HLHA beads. They suggest that the high-alumina sand would have come from the abundant sillimanite-rich and mica schist deposits in Ile-Ife. As they explain, “Such high-lime, high-alumina glass has been found only in West Africa . . . and is not known from Europe, the Middle East or Asia, ruling out the possibility that the glass was imported” (Lankton et al., 2006, p. 111). The production, they suggest, centered in or near Ile-Ife, providing the first strong evidence for early primary glass production in sub-Saharan Africa. I. C. Freestone (2006), in a commentary on Lankton et al.’s paper published in the same issue of the Journal of African Archaeology, noted the stunningly unique chemical composition of Ife glass. He speculated that the high metaluminous content of the glass samples (alumina and alkalis being roughly equal) was possibly achieved through the addition of supplementary alkali feldspar (immature granitic sand most likely from local ponds or rivers) and lime-rich matter, the latter perhaps incorporating snail shells, which are nearly pure calcium carbonate. According to Freestone (2006), “ . . . we might have here a truly African technology where the raw materials were added as rock or mineral, something without precedent in early glass making as we currently understand it” (p. 140). Our article presents new sets of evidence that are based on the use of a different instrumental analysis and samples from a different archaeological site in the region. The result provides affirmative answers that there was indeed a unique glassmaking tradition among the ancestral Yoruba.

Archaeology and Geology of Early Osogbo

Recent archaeological research in Osun Grove (Figure 2), a Nigerian national monument and UNESCO World Heritage Site, has yielded new evidence of glass cullet, fragments of a glass-lined crucible, and glass beads. This is the first time that glass cullet is reported outside Ile-Ife in the Yoruba region. The excavations that produced these materials took place in an abandoned settlement, Early Osogbo, in the northeast corner of Osun Grove, 45 km north of Ile-Ife. The drainage pattern of the area is moderately dense and dendritic, dominated by the Osun River and its tributaries, which are largely controlled by the structural trends within the Basement Complex terrain. Geologically, the study area lies largely within the Precambrian Basement Complex of southwestern Nigeria, and belongs to the Pan-African mobile belt east of West African Craton. The major rock groups in the study area are migmatites complex (including banded and augen gneisses as well as pegmatites) and metasediments (consisting of schist, quartzites and amphibolites). The dominant basement rocks in the Osogbo area are schist and granite pegmatites, associated with quartzite ridges forming the characteristic undulating terrain. A significant feature of the Osun Grove geology is the presence of very thick, granite-derived grayish sand, which must have accumulated in the Osun floodplain for thousands of years. The stratigraphic profiles, chronometric dates, and oral traditions place the occupation of Early Osogbo settlement at the period between approximately 1575 and 1700 (for details, see Ogundiran, 2014). A diverse range of artifacts were found in Osun Grove. The ones that are relevant to the substance of this article are glass cullet, glass beads, glass waste, and a fragment of glass-lined crucible (Figure 3).

Figure 2.

Osun Grove, Osogbo.

Figure 3.

(A) Glass-lined crucible (interior), (B) glass cullet, and (C) beads.

Analysis and Results

We have undertaken the study of chemical compositions of the excavated glass cullet and beads in Early Osogbo using the Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS). The goal of the study is to understand the chemical components of the glass artifacts with the view to making inferences about the raw materials used in the glass-manufacturing processes in Early Osogbo, especially in light of previous studies conducted on glass and glass bead manufacture in Ile-Ife. The materials analyzed in this study consist of 23 glass cullet samples (unshaped glass objects), 18 glass beads and bead wasters, a glass-lined crucible fragment, three samples of pegmatites, and two samples of snail shells all from Early Osogbo. These represent the largest number of excavated glass samples and associated materials from a single site in Yorubaland ever analyzed for geochemical signatures.

LA-ICP-MS analysis offers the most detailed chemical properties of any inorganic object. The method is non-destructive, and it can identify the four types of minerals that an object is made of: major, minor, trace, and rare earth (Dussubieux and Gratuze 2003). The major elements are the bulk substances that constitute an object. These are elements that define the basic character of the material in question. They account for the majority of the weight of the object and are measured in percentages. For example, silica (sand oxide) is the basic component of any glass. The minor elements are the secondary properties of an object and these generally account for less than 1% of the object. Trace elements occur in very small concentrations, less than 0.1% of the object weight, and are not fundamental to defining the chemical essence of an object. But trace elements tend to determine the color of a glass object, and they are particularly useful for determining the source and environment from which an object originated. The fourth, rare earth elements (REEs), refers to elements that tend to occur together in nature, in the same ore deposits. Although REEs are abundant, they are rarely present together in viable commercial quantity, hence their name “rare earth”. In glass identification, REEs serve purposes that are similar to those of trace elements. They may contribute to the degree of luster, hardness, opacity, and color variation in an object. Both trace and REEs are measured in 1,000 parts per million because they are less than 0.1% of an object’s weight.

Major Element Compositions of Glass Cullet and Beads

The major element compositions of both glass cullet and beads are presented in Table 1. Major oxides in the glasses and glass beads generally comprise of silica (SiO₂), sodium (Na₂O), magnesium (MgO), calcium (CaO), aluminum (Al₂O₃), and potassium (K₂O). Minor elements include manganese (MnO), iron (Fe₂O₃), phosphate (P₂O₅), lead (PbO₂), and copper (CuO). The review of the data in Table 1 shows that the overwhelming majority of the cullet are of the high-alumina, high-calcium type. The Na₂O contents are more variable, 1.12% to 4.37%, while the alumina content is between 11.31% and 19.74%, and calcium varies between 9.98% and 16.78%. The CaO-to-alumina ratio is about 1 in most samples. The contents of magnesium are generally lower than 0.1wt%, while the potassium oxides vary between 2.3% and 6.74%, indicating the use of feldspathic minerals as both fluxes and stabilizers. P₂O₅ is generally low, mostly less than 0.3%, which indicates that plant ash was not utilized in the making of these glass artifacts (Freestone, 2006).

Table 1.

Major and Minor Oxides of Glass Artifacts, Early Osogbo.

Sample Number	Type	SiO₂	Na₂O	MgO	Al₂O₃	P₂O₅	K₂O	CaO	MnO	Fe₂O₃	CuO	SnO₂	PbO₂
OSG-G-1	Cullet	61.1	2.2	0.1	13.7	0.08	7.1	14.6	0.6	0.5	0.01	0.00	0.00
OSG-G-3	Cullet	58.9	3.7	0.1	16.1	0.12	3.2	16.4	0.2	1.3	0.00	0.00	0.00
OSG-G-4	Cullet	62.5	4.3	0.1	15.3	0.10	2.3	11.7	0.1	3.5	0.01	0.00	0.01
OSG-G-5	Cullet	64.6	2.3	0.1	13.9	0.06	3.3	14.2	0.8	0.6	0.02	0.00	0.01
OSG-G-6	Cullet	63.1	4.1	0.1	14.2	0.09	3.0	13.9	0.6	0.8	0.01	0.00	0.01
OSG-G-8	Cullet	62.8	2.0	0.1	12.3	0.10	7.7	14.4	0.2	0.5	0.00	0.00	0.01
OSG-G-9	Cullet	62.5	3.5	0.0	14.1	0.06	4.9	13.6	0.8	0.4	0.02	0.00	0.01
OSG-G-10	Cullet	63.7	4.5	0.0	13.4	0.04	2.6	14.7	0.5	0.5	0.01	0.00	0.01
OSG-G-11	Cullet	65.7	3.4	0.0	13.6	0.08	4.2	11.6	0.8	0.4	0.02	0.00	0.00
OSG-G-12	Cullet	61.1	4.5	0.0	14.5	0.07	3.8	14.7	0.6	0.5	0.01	0.00	0.01
OSG-G-13	Cullet	64.5	4.4	0.0	13.6	0.07	3.1	13.6	0.4	0.4	0.01	0.00	0.01
OSG-G-14	Cullet	67.0	3.8	0.0	12.6	0.07	3.2	12.4	0.6	0.3	0.01	0.00	0.01
OSG-G-15	Cullet	65.7	1.9	0.1	11.3	0.10	7.9	12.0	0.5	0.4	0.01	0.00	0.00
OSG-G-16	Cullet	62.4	4.3	0.0	14.0	0.08	3.2	14.8	0.7	0.3	0.01	0.00	0.00
OSG-G-17	Cullet	66.6	5.7	0.1	12.8	0.12	0.9	12.3	0.7	0.7	0.02	0.00	0.01
OSG-G-18	Cullet	62.0	4.1	0.0	13.7	0.07	5.2	13.8	0.1	0.9	0.00	0.00	0.02
OSG-G-19	Cullet	62.0	3.1	0.1	13.0	0.09	4.0	15.2	0.1	2.4	0.01	0.00	0.01
OSG-G-20	Cullet	61.1	1.9	0.1	13.5	0.14	9.0	13.7	0.3	0.4	0.00	0.00	0.01
OSG-G-22	Cullet	58.7	0.9	0.3	19.8	0.29	6.7	10.0	0.3	3.0	0.01	0.00	0.01
OSG-G-23	Cullet	61.3	2.9	0.4	13.1	0.25	6.5	14.8	0.1	0.7	0.00	0.00	0.01
OSG-G-24	Cullet	62.8	1.9	0.2	12.6	0.07	6.7	14.7	0.4	0.6	0.01	0.00	0.01
OSG-G-47	Cullet	54.2	2.1	1.4	15.4	0.59	6.5	19.0	0.1	0.8	0.00	0.00	0.00
OSG-G-49	Cullet	57.1	1.1	0.3	17.8	0.24	6.6	14.2	0.4	2.1	0.01	0.00	0.02
OSG-G-28	Bead	60.9	5.8	0.1	14.9	0.10	1.2	15.8	0.8	0.4	0.02	0.00	0.00
OSG-G-29	Bead	62.4	4.3	0.0	14.0	0.08	3.2	14.8	0.7	0.3	0.01	0.00	0.00
OSG-G-30	Bead	61.5	2.3	0.0	14.3	0.12	8.5	12.6	0.3	0.4	0.02	0.00	0.00
OSG-G-31	Bead	61.1	2.6	0.2	13.6	0.13	6.4	15.0	0.4	0.6	0.00	0.00	0.00
OSG-G-33	Bead	62.1	3.6	0.1	14.1	0.09	5.1	13.6	0.4	0.8	0.01	0.00	0.00
OSG-G-34	Bead	61.6	4.2	0.0	14.1	0.06	3.4	14.8	0.7	0.9	0.02	0.00	0.00
OSG-G-35	Bead	60.6	3.2	0.1	14.5	0.08	5.7	14.1	0.6	1.1	0.01	0.00	0.01
OSG-G-36	Bead	62.9	3.7	0.1	13.3	0.07	4.1	14.6	0.7	0.4	0.02	0.00	0.01
OSG-G-37	Bead	61.8	3.6	0.0	13.8	0.07	4.9	14.8	0.6	0.3	0.02	0.00	0.01
OSG-G-38	Bead	64.5	4.1	0.0	13.5	0.09	5.3	12.1	0.1	0.3	0.00	0.00	0.00
OSG-G-39	Bead	69.1	5.8	0.6	12.7	0.16	3.7	4.5	0.3	2.6	0.21	0.01	0.04
OSG-G-40	Bead	62.9	4.1	0.0	14.1	0.06	3.6	14.3	0.6	0.4	0.01	0.00	0.01
OSG-G-41	Bead	60.4	5.8	0.1	14.9	0.09	1.2	16.1	0.8	0.4	0.02	0.00	0.00
OSG-G-42	Bead	61.2	3.6	0.1	13.9	0.07	5.0	14.9	0.6	0.5	0.01	0.00	0.01
OSG-G-43	Bead	60.0	3.9	0.0	15.0	0.09	4.7	14.9	0.8	0.4	0.03	0.00	0.01
OSG-G-44	Bead	63.5	2.4	0.0	13.5	0.13	7.9	12.0	0.3	0.3	0.01	0.00	0.00
OSG-G-45	Bead	64.2	3.4	0.0	13.0	0.07	4.3	13.5	1.0	0.3	0.02	0.00	0.01
OSG-G-46	Bead	64.1	2.5	0.0	13.4	0.09	5.6	12.9	0.8	0.4	0.02	0.00	0.00

Note. Elemental oxides: SiO₂ = silica; Na₂O = sodium; MgO = magnesium; Al₂O₃ = aluminium; P₂O₅ = phosphorous; K₂O = potassium; CaO = calcium; MnO = manganese; Fe₂O₃ = iron; CuO = copper; PbO₂ = lead; SnO₂ = tin.

The beads present similar compositions to the glass cullet (discussed above) in terms of the major elements SiO₂, CaO, Al₂O₃, K₂O, and Na₂O. The beads have very low P₂O₅ (<0.5) value, and all have less than 0.5% MgO with the majority less than 0.1%, clearly ruling out plant ash raw material for the manufacture of the glass beads (as plant ash contains high P₂O₅ and MgO). Given the fact that the glass of the Islamic World (including India) were characterized by plant ash while those of the Medieval Europe were based on wood-ash glass, the geochemical signatures of the Early Osogbo glass—HLHA—were unique and have been found only in Yorubaland. But this is not all; there are other lines of geochemical and archaeological evidence showing that this HLHA glass derived from local minerals and geological deposits.

Trace Element and Rare Earth Content of Glass Cullet and Beads

The cullet and glass are also broadly similar in trace element composition (Table 2). Any variation here would be due to different sources of raw materials or maturity levels of the sediments (silica). The following trace elements are preponderant in the chemical composition of the Early Osogbo glass: lithium (Li), beryllium (Be), boron (B), scandium (Sc), titanium (Ti), vanadium (V), chromium (Cr), nickel (Ni), cobalt (Co), zinc (Zn), rubidium (Rb), strontium (Sr), zirconium (Zr), niobium (Nb), antimony (Sb), cesium (Cs), and barium (Ba). Arsenic (As) element also occurs but in very small quantity. The trace elements—titanium, zirconium, vanadium, and chromium—are related to the heavy minerals and mafic fractions (e.g., rutile, zircon, garnet, biotites), both of which are common accessory minerals within granite and pegmatites. The base metal components (Ni, Cr, Cu, Ti) are exotic admixtures from nearby mafic-ultramafic assemblages, which are ubiquitous in the erosional path of the Osun River and its tributaries. The manganese (0.1%-0.8%) and cobalt (100-600 parts per million [ppm]) contents of the glass are significantly high. These could be natural impurities present in the deposits from which the silica raw materials were mined. And, these impurities may be the very reason why such sources of raw materials were preferred. It is also likely, however, that cobalt specimens were deliberately added to the glass-manufacturing process.

Table 2.

Trace Elements in the Glass Artifacts, Early Osogbo.

Sample Number	Li	Be	B	Sc	Ti	V	Cr	Ni	Co	Zn	As	Rb	Sr	Zr	Nb	Sb	Cs	Ba
Cullet
OSG-G-1	53	1.6	8	1.7	148	17	6	123	482	88	0.0	297	314	10	2.1	0.0	6.7	175
OSG-G-3	27	85.5	52	2.5	257	18	15	25	130	71	0.0	349	270	25	34.7	0.2	8.6	172
OSG-G-4	14	3.0	14	3.4	501	12	37	47	45	86	0.0	99	223	31	14.6	0.4	2.9	134
OSG-G-5	37	1.5	15	1.8	87	11	4	91	934	101	0.0	127	264	20	2.8	0.6	2.4	180
OSG-G-6	50	1.8	16	2.5	275	24	16	89	639	113	0.0	107	276	18	4.1	0.2	1.3	188
OSG-G-8	16	6.3	9	1.3	74	5	51	100	175	15	0.0	354	279	19	6.2	0.2	6.4	228
OSG-G-9	67	1.7	12	1.7	117	11	2	103	793	96	0.0	187	266	23	2.2	0.2	2.8	225
OSG-G-10	41	1.3	10	1.6	90	10	4	62	542	66	0.0	84	297	16	3.5	0.3	1.3	184
OSG-G-11	56	2.0	18	2.1	99	14	5	102	769	95	0.0	137	187	13	1.9	0.2	1.6	167
OSG-G-12	46	2.2	8	2.1	156	17	7	67	729	76	0.0	129	272	10	2.7	0.2	1.8	200
OSG-G-13	29	2.3	16	1.4	94	11	5	54	290	46	0.0	104	270	12	2.2	0.4	1.3	158
OSG-G-14	45	2.2	15	1.4	73	10	2	60	690	66	0.0	96	193	14	1.9	0.1	1.8	154
OSG-G-15	18	1.4	8	1.6	73	6	69	306	528	20	0.0	271	320	16	1.0	0.1	1.9	593
OSG-G-16	35	1.6	12	1.4	98	10	3	46	143	39	0.0	341	185	11	2.7	0.0	6.2	110
OSG-G-17	54	9.0	42	2.0	156	17	7	95	375	81	57.0	46	289	17	4.0	1.0	4.0	239
OSG-G-18	11	2.0	9	1.0	108	3	4	3	2	12	0.0	178	314	15	3.0	1.0	3.0	143
OSG-G-19	40	16.0	35	2.0	313	10	11	8	25	21	3.0	207	289	38	10.0	0.0	5.0	125
OSG-G-20	14	1.0	9	1.0	90	4	45	96	306	12	1.0	249	267	29	1.0	0.0	4.0	226
OSG-G-22	18	2.0	15	7.0	3,239	37	46	107	204	121	25.0	209	303	34	20.0	0.0	5.0	976
OSG-G-23	16	36.0	7	1.0	273	7	9	8	9	26	0.0	358	360	66	26.0	0.0	7.0	204
OSG-G-24	24	1.2	9	2.6	148	7	64	184	368	23	0.0	386	273	26	17.5	0.2	8.5	230
OSG-G-47	29	4.0	14	2.0	920	16	8	9	9	5	0.0	247	800	97	11.0	0.0	4.0	612
OSG-G-49	17	2.0	14	5.0	1,585	24	49	179	302	89	0.0	201	385	44	15.0	0.0	4.0	936
Beads
OSG-G-28	69	10.9	41	2.1	107	20	4	123	587	97	0.0	58	326	18	4.2	0.3	2.8	347
OSG-G-29	58	2.1	8	1.3	190	12	3	85	906	88	0.0	128	235	23	2.6	0.0	2.4	160
OSG-G-30	37	1.5	6	1.4	182	10	7	45	176	34	0.0	386	248	23	2.5	0.0	9.1	132
OSG-G-31	27	8.2	9	1.8	195	7	47	157	266	26	0.0	362	364	29	14.6	0.5	4.8	287
OSG-G-33	52	1.5	7	2.7	256	14	9	79	502	52	0.0	192	221	27	4.0	0.0	2.5	151
OSG-G-34	58	2.3	8	3.1	190	16	11	120	587	81	0.0	111	298	26	2.8	0.0	1.6	328
OSG-G-35	47	1.9	8	2.8	225	15	15	58	639	52	0.0	207	287	21	3.1	0.0	3.2	237
OSG-G-36	56	2.3	16	2.0	147	13	1	103	565	87	0.0	143	203	25	3.0	0.3	2.4	241
OSG-G-37	51	1.6	13	1.4	112	13	1	73	652	74	0.0	170	252	13	2.1	0.2	2.6	256
OSG-G-38	11	2.0	8	0.8	90	3	3	2	1	5	0.0	228	220	6	5.1	0.0	3.6	105
OSG-G-39	28	1.7	42	1.9	184	10	9	35	214	275	24.5	115	109	25	2.1	6.6	1.9	86
OSG-G-40	49	1.8	12	1.6	124	15	6	76	546	75	0.0	126	278	15	2.9	0.3	1.3	169
OSG-G-41	66	11.0	41	2.0	106	20	4	123	574	99	0.0	57	320	18	4.1	0.3	2.7	339
OSG-G-42	47	1.9	13	1.4	157	14	1	76	674	82	0.0	178	256	14	2.5	0.3	2.9	265
OSG-G-43	93	1.9	10	1.5	90	16	1	145	898	125	0.0	190	296	19	3.0	0.2	2.9	184
OSG-G-44	117	5.8	297	3.1	145	17	2	126	906	135	0.0	391	224	22	27.6	0.2	8.2	195
OSG-G-45	77	1.8	22	1.8	126	20	2	156	913	154	0.0	149	258	15	2.3	0.7	2.1	279
OSG-G-46	54	6.0	56	1.9	120	22	6	130	747	116	0.0	352	180	14	3.3	0.2	9.7	131

Note. Li = lithium; Be = beryllium; B = boron; Sc = scandium; Ti = titanium; V = vanadium; Cr = chromium; Ni = nickel; Co = cobalt; Zn = zinc; As = arsenic; Rb = rubidium; Sr = strontium; Zr = zirconium; Nb = niobium; Sb = antimony; Cs = cesium; Ba = barium.

The trace elements of the glass cullet and beads (Table 2) show great variation and elevated levels of Mn, Fe, Cu, Pb, and Sn, which are controlled by the mineralogy of granite, pegmatite, and quartzites found in the area. Sr contents of the glasses vary between 107 and 368 ppm, while rubidium ranges between 87 and 546 ppm. These are elemental signatures of sand derived from a granitic protolith. There are also elevated levels of copper and tin at 0.1% to 1.35%, and these would have served as colorants for both the glass cullet and beads.

As a further step to explore the different sources of certain elements in the glass and glass beads, we evaluated the REE composition of the glasses. The ones that concern us here, because of their frequency of occurrence, are the lanthanide series of chemical elements: lanthanum (La), cerium (Ce), neodymium (Nd), samarium (Sm), europium (Eu), gadolinium (Gd), terbium (Tb), dysprosium (Dy), holmium (Ho), erbium (Er), thulium (Tm), ytterbium (Yb), and lutetium (Lu). These REEs are widely acknowledged as the backbone of modern industries from the production of eyeglasses and cell phones to the manufacture of hybrid cars and aircrafts. The distribution patterns of REEs in both the glass cullet and beads are presented in Table 3. These elements would have been present naturally in the bulk recipe, such as pegmatite, used for glass manufacture. The values of these REEs show that the raw materials used in the manufacture of the Early Osogbo glass are weathering products of granitic rocks, which are abundant in the vicinity of Early Osogbo. These granitic sands can be classified as immature sand derived from weathering of granite and quartzites. Heavy minerals, such as monazite and zircon, may also be present to contribute to these rare earth patterns.

Table 3.

Rare Earth Elements in the Glass Artifacts, Early Osogbo.

	La	Ce	Nd	Sm	Eu	Gd	Tb	Dy	Ho	Er	Tm	Yb	Lu
Cullet
OSG-G-1	3.6	36.4	2.8	0.7	0.2	0.6	0.1	0.6	0.1	0.3	0.1	0.4	0.1
OSG-G-3	17.1	50.3	10.8	1.7	0.4	1.1	0.2	0.9	0.2	0.5	0.1	0.6	0.1
OSG-G-4	10.8	21.9	8.2	2.2	0.3	1.3	0.2	1.3	0.3	0.7	0.1	0.7	0.1
OSG-G-5	6.0	22.1	6.0	1.4	0.4	1.2	0.2	1.4	0.3	0.7	0.1	0.9	0.1
OSG-G-6	16.2	74.5	6.7	1.2	0.3	1.0	0.2	1.0	0.2	0.6	0.1	0.9	0.1
OSG-G-8	4.6	9.1	3.9	1.0	0.2	0.7	0.1	0.7	0.1	0.4	0.1	0.4	0.1
OSG-G-9	5.3	26.9	4.9	1.1	0.3	1.0	0.2	1.1	0.2	1.6	0.1	0.8	0.1
OSG-G-10	8.4	28.9	6.4	1.5	0.4	1.1	0.2	1.3	0.2	0.6	0.1	0.8	0.1
OSG-G-11	10.1	33.8	4.9	0.9	0.2	0.7	0.1	0.7	0.1	0.4	0.1	0.5	0.1
OSG-G-12	108.6	82.1	30.5	3.9	0.9	2.4	0.3	1.7	0.3	0.7	0.1	0.8	0.1
OSG-G-13	3.9	43.0	2.8	0.7	0.2	0.7	0.1	0.7	0.1	0.4	0.1	0.6	0.1
OSG-G-14	5.3	20.5	5.0	1.1	0.3	0.8	0.1	0.8	0.1	0.4	0.1	0.5	0.1
OSG-G-15	5.8	14.9	5.3	1.1	0.4	0.8	0.1	0.6	0.2	0.2	0.0	0.3	0.0
OSG-G-16	7.4	23.5	6.3	1.4	0.3	1.1	0.2	1.3	0.2	0.6	0.1	0.8	0.1
OSG-G-17	7.0	24.0	10.0	3.0	1.0	2.0	0.0	2.0	0.0	1.0	0.0	1.0	0.0
OSG-G-18	5.0	9.0	3.0	1.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	1.0	0.0
OSG-G-19	5.0	11.0	5.0	1.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0
OSG-G-20	4.0	14.0	3.0	1.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
OSG-G-22	44.0	106.0	34.0	6.0	1.0	5.0	1.0	5.0	1.0	3.0	0.0	3.0	0.0
OSG-G-23	5.0	10.0	4.0	1.0	0.0	1.0	0.0	1.0	0.0	0.0	0.0	0.0	0.0
OSG-G-24	5.8	18.1	5.8	1.4	0.3	1.0	0.2	0.9	0.2	0.5	0.1	0.5	0.1
OSG-G-47	20.0	27.0	14.0	3.0	1.0	2.0	0.0	2.0	0.0	1.0	0.0	1.0	0.0
OSG-G-49	40.0	104.0	31.0	6.1	1.0	5.0	1.0	4.0	1.0	3.0	0.0	3.0	0.0
Beads
OSG-G-28	13.5	26.6	17.4	4.4	1.2	4.1	0.6	3.7	0.7	1.7	0.3	1.9	0.3
OSG-G-29	7.4	23.5	6.3	1.4	0.3	1.1	0.2	1.3	0.2	0.6	0.1	0.8	0.1
OSG-G-30	5.5	12.5	6.9	1.6	0.4	1.2	0.2	1.1	0.2	0.5	0.1	0.5	0.1
OSG-G-31	4.5	13.9	3.9	1.0	0.2	0.7	0.1	0.8	0.1	0.3	0.1	0.4	0.1
OSG-G-33	67.6	20.8	99.5	23.0	6.5	18.9	2.8	15.9	2.9	7.3	1.0	6.4	0.9
OSG-G-34	7.9	18.8	6.5	1.6	0.4	1.5	0.3	1.8	0.3	1.0	0.2	1.2	0.2
OSG-G-35	44.6	92.9	15.7	2.2	0.6	1.5	0.2	1.3	0.2	0.7	0.1	0.7	0.1
OSG-G-36	6.5	20.9	6.4	1.5	0.4	1.3	0.2	1.5	0.3	0.8	0.1	1.0	0.1
OSG-G-37	11.5	45.8	5.9	1.2	0.3	0.9	0.1	0.7	0.1	0.4	0.1	0.5	0.1
OSG-G-38	2.1	4.2	1.7	0.5	0.1	0.4	0.1	0.3	0.1	0.1	0.0	0.2	0.0
OSG-G-39	6.0	14.9	4.1	1.0	0.2	0.8	0.1	0.8	0.2	0.4	0.1	0.5	0.1
OSG-G-40	20.9	56.7	7.3	1.3	0.3	0.9	0.2	1.0	0.2	0.6	0.1	0.7	0.1
OSG-G-41	13.2	25.1	16.9	4.3	1.2	3.9	0.6	3.5	0.6	1.8	0.3	1.9	0.3
OSG-G-42	12.3	49.6	6.4	1.3	0.3	0.9	0.1	1.0	0.2	0.5	0.1	0.6	0.1
OSG-G-43	23.9	33.7	66.2	15.8	3.6	10.4	1.2	5.0	0.7	1.5	0.2	1.3	0.2
OSG-G-44	2.8	7.0	2.2	0.6	0.1	0.5	0.1	0.8	0.1	0.4	0.1	0.5	0.1
OSG-G-45	16.5	46.7	8.9	1.7	0.5	1.2	0.2	1.2	0.2	0.6	0.1	0.8	0.1
OSG-G-46	11.7	53.3	6.4	1.1	0.3	0.8	0.1	0.7	0.1	0.3	0.1	0.4	0.1

Note. La = lanthanum; Ce = cerium; Nd = neodymium; Sm = samarium; Eu = europium; Gd = gadolinium; Tb = terbium; Dy = dysprosium; Ho = holmium; Er = erbium; Tm = thulium; Yb = ytterbium; Lu = lutetium.

Chemical Composition of the Likely Raw Materials

A significant feature of the Early Osogbo archaeological deposit is the excavation of a silo of pegmatite sand that could have been used for the manufacture of the excavated glass (Figure 4). The deposit consisted of pebbles of two types of feldspar pegmatite in a matrix of fine-grain pegmatite. One feldspar type is reddish and is identified as andesine, while the other is whitish and is identified as orthoclase, sandwiched between nearly colorless quartz crystals (Figure 4). This is surrounded by a veneer of clay, which we identified by X-ray to be mostly kaolinite with minor chlorite. In Table 4, we present the analytical data of the major chemical elements in three pegmatite samples together with data on two shells found in the excavation.

Figure 4.

Pegmatite storage pit (during excavation) and pegmatite samples.

Table 4.

Major Elements of Three Pegmatite Samples (OSG-S-1, OSG-S-2, and OSG-S-3) and Two Snail Shells, Early Osogbo.

	OSG-S-1	OSG-S-2	OSG-S-3	Shell7_1	Shell8_1
SiO₂	54.45	54.76	61.58	0.47	0.09
Na₂O	0.24	0.42	0.07	0.02	0.15
MgO	6.42	4.21	1.04	0.06	0.02
Al₂O₃	28.39	30.44	26.83	0.23	0.02
P₂O₅	0.04	0.06	0.1	0.1	0.02
K₂O	1.21	1.92	1.04	0.04	0.05
CaO	0.62	0.6	0.53	98.71	99.17
MnO	0.07	0.058	0.13	0.01	0.01
Fe₂O₃	8.47	7.39	8.46	0.36	0.42
CuO	0.01	0.01	0.01	0.001	0.001
PbO₂	0.001	0.001	0.01	0	0

Note. Elemental oxides: SiO₂ = silica; Na₂O = sodium; MgO = magnesium; Al₂O₃ = aluminum; P₂O₅ = phosphorous K₂O = potassium; CaO = calcium; MnO = manganese; Fe₂O₃ = iron; CuO = copper; PbO₂ = lead.

The pegmatite samples are similar in mineralogy to the recipe used in an experimental melting of sand aimed at reconstructing the process of ancient glass-making among the Yoruba. The melts produced in the experiment were analyzed with the electron-probe facility at the University of Georgia (Ige, 2009). For this investigation, four recipes that represent possible raw materials for the much earlier Ife glass factory (ca. 11th-15th centuries) were selected for analysis and subsequent glassmaking. By varying the proportions of quartz, feldspar and mica, different raw material mixtures were obtained. The chemical composition of the recipes was obtained using an energy-dispersive XRF. The approximate mineralogical compositions of raw materials are as follows:

Recipe 1 : 1 \frac{1}{2} g quartz + 3 g oligoclase + 1 \frac{1}{2} g biotite .

Recipe 2 : 1 \frac{1}{2} g quartz + 2 g oligoclase + 1 g microcline + \frac{1}{2} g biotite + \frac{1}{2} g muscovite .

Recipe 3 : 2 g quartz + 2 g oligoclase + 1 g microcline+ \frac{1}{2} g biotite + \frac{1}{2} g muscovite .

Recipe 4 : 2 g quartz + 2 g oligoclase + \frac{1}{2} g biotite .

Powered samples of the recipes were melted in silica crucibles placed in a Wildbar furnace at a temperature of 1050 °C to 1150 °C for times ranging between 15 and 48 hours. The furnace is equipped with a chromel-alumina thermocouple in order to obtain approximate maximum temperature reading.

The recipes were not placed directly in the furnace but were first inserted in a horizontal quartz tube and then put into the furnace. Due to the furnace capacities, a total mass of only 4 to 6 g per experiment was possible. Sometimes the samples were sintered prior to the melting experiment. After cooling, the frit was thoroughly ground and homogenized in a boron carbide pistil mill in order to get rid of gas bubbles prior to melting at temperatures of 1150 °C to 1300 °C. Glass compositions and micrographs from melting experiments were obtained with the assistance of Dr. Sam Swanson using the electron-probe facility of the Geology Department of the University of Georgia. The detail of the experiment has been presented in Ige (2009). This limited but useful experiment indicates that the glass-making artisans in Yorubaland must be using the abundant granitic pegmatite resources in their environment to manufacture glass. The current archaeological evidence and geochemical analysis of the processed pegmatite deposits in Early Osogbo, along with the analysis of the glass artifacts, show that granitic pegmatite was indeed used in glass manufacture in Yorubaland.

Discussion

These LA-ICP-MS results indicate that the major element composition of the glass beads and cullet consist of five oxides—SiO₂, Na₂O, Al₂O₃, K₂O, and CaO—constituting about 98% of the materials. MgO, P₂O₅, MnO, Fe₂O₃, and CuO constitute about 1.5%. Cu, Pb, and Co, account for the lowest weight percentage, and would have been used as colorants for the glass beads. The geochemical data of Early Osogbo glass beads, the glass-lined crucible fragment, and cullet have broadly similar chemical characteristics in terms of their low MgO, high CaO and Al₂O₃, and extensive enrichment in REE patterns. One of the most intriguing finds is the discovery of a silo containing processed pegmatite (glass sand), the kind of raw materials that would have been used for the glass manufacture. The pegmatite derived from the thick granitic sands in the floodplain of the Osun River, and these could have been melted together with local colorants for the manufacture of the glass beads. Cobalt and MnO contents are especially high in many of the samples, and are likely derived from the ultramafic assemblage close to the granite body that is transported as erosional products from nearby Ilesa greenstone belts. The high levels of Rb and Sr in the assemblage point to derivation from the granitic rock.

The high concentration of CaO (calcium) would have resulted from two possibilities: either the melting sand, through firing, flushed out the sodium and caused the elevation of CaO, or that CaO (in powdered form) was intentionally added to the recipe. The possible source of the high-calcium content in the Osogbo glass is perhaps the ubiquitous snail shells. One possibility is that ground snail shells were added to the basic glass composition in powdery form. In a glassmaking experimental study, Ige (2009, 2010) reported that most of the drawn glass beads from Ile-Ife were manufactured from glass produced by melting granitic sands, sometimes blended with snail shells. Our analytical data in Table 4 support this.

The high-alumina contents and variable composition of the Osun Grove glass cullet and glass beads are also consistent with the geological raw materials present in the Osogbo area. Osun Grove lies in a floodplain of the culturally famous Osun River. Over several thousands of years, thick white sand derived from voluminous quartz-rich rocks such as granites and pegmatites have been deposited on the banks of the river. Thesewhite sand deposits were likely mined in the past for pottery and glass manufacture. Some pockets of rocks of basic composition are also found in the area as an extension of the ultramafic assemblage in the nearby Ilesa greenstone belt. These basic rocks are sources of the iron-encrusted lateritic soil type, which are composed of iron minerals such as goethite, limonite, and disseminated iron oxides (e.g., ilmenite, magnetite, and rutile). Such variety of rock units will give rise to sand of complex composition. For example, the granitic rock can be sources for sodium, potassium, alumina, and silica because of the preponderance of sodic and potash feldspars as well as quartz in their mineralogy; while magnesium, iron, and manganese contents could be derived from thick lateritic soils overlying weathered basic rocks. These rocks are also extensively rich in base metals such as nickel, chromium, and cobalt, which have been weathered and deposited as sand (Ige, Durotoye, & Oluyemi, 2005). It has already been reported that iron masters in the Ile-Ife area exploited a variant of such sand in making high-quality steel (Ige & Rehren, 2003). And, this extends to glass. As Lankton et al. (2006) note,

In or near Ile-Ife the ready availability of local raw materials, the tradition of high-temperature technology in the form of copper alloy or iron working and local and regional demand for glass beads combined to stimulate primary and secondary glass technologies that would continue for many centuries. (p. 133)

Cobalt-bearing mafic and ultramafic rocks have been found in several localities around Osogbo, especially Mokuro-Ife and Isaobi-Ilesa, where incontrovertible evidence of sub-economic accumulation of base metals like cobalt, titanium, and nickel have been detected and may have been mined in antiquity as sources of cobalt colorants for bead manufacture in Yorubaland. The evidence of local concentration of these metals is overwhelming (Ajayi & Suh, 1999; Emofurieta, Aladesewa, & Ogunseji, 1995).

Conclusion

The LA-ICP-MS analysis of excavated glass cullet, glass beads, a crucible fragment, and processed pegmatite in Early Osogbo (Osun Grove) provide compelling results. The chemical signatures of these samples are grossly homogeneous and characterized by HLHA components. These are in turn similar to the ones previously documented in Ile-Ife by Lankton et al. (2006) using SEM-EDS, EPMA, and XRF methods. The Early Osogbo glasses in particular are synonymous with a glass fabricated from the melting of immature granite-derived sediments in an experiment carried out by the second author (Ige, 2009). Our work also provides empirical evidence in support of Ian Freestone’s previous hypothesis indicating that Yoruba glassmakers were adding nearly pure calcium carbonate, derived from immature granitic sand in river sources and lime-rich matter, to their glassmaking raw material recipe. This “truly African technology . . . without precedent in early glass making as we currently understand it,” according to Freestone (2006, p. 140), is now demonstrated in the Early Osogbo archaeological contexts and in the geochemical evidence of the glass assemblage. Our findings not only present evidence for the unique chemical signatures of Yoruba glass but also identify the local raw materials used for this complex technology.

Most important, our results show that the Early Osogbo glasses were granite-based, different from the plant ash–based glass of the Islamic World (including India), Late Bronze Age Egypt, and Mesopotamia; the mineral natron glass of the Greek, Roman, and Byzantine Empires; the lead and lead-barium glass of Han China; and the wood-ash and ash-lime glass of Medieval Europe (Rehren & Freestone, 2015). The trace element contents and rare earth patterns, for the first time, offer new perspectives on the likely raw materials for the Early Osogbo glass. These LA-ICP-MS results are significant. They challenge us to reconsider the history of glass technology in Africa, south of the Sahara, especially as late as the 17th century. These geochemical results also open up new fodders for rethinking the relationships between glass technology and the Yoruba political, cultural, economic, ideological, and social practices especially between 1000 and 1800 ad. So, the Late Omotoso Eluyemi was right that “our ancestors were material scientists.” Through experimental initiatives, the Yoruba ancestors attained the expertise that enabled them to develop a unique recipe for glass manufacture. They mined and processed feldspathic pegmatite (granites), cobalt, titanium, rubidium, and nickel, among others, and combined these with powdered snail shells. The ancestral Yoruba material scientists were fully aware of the elemental properties of these materials. They then transformed this recipe, through precise pyro-technology (firing) techniques, to produce glass and glass beads that served as the currency for negotiating political power, economic relations, and cultural/spiritual values. The legacies of this technology and the beaded artworks they inspired continue to have relevance in the Yoruba world, both in Africa and the African Diaspora (Drewal & Mason, 1998).

We hope this article serves as an inspiration for the kind of collaborative work that needed to be done in Black Studies between the humanists and scientists about the history of technology in Africa. Material science has played an important role in the exploitative experience that African peoples have suffered in the rush by external agents for the continent’s raw materials to power the engine of modern industrialization, both in the West and East. In order for Black Studies scholars to effectively advocate for the preservation of African resources for the benefit of Africans, it is imperative that they know about how African ancestors made use of material science in the past. Africa’s indigenous knowledge must be part of the discussion of how Africa-descended populations can use the continent’s resources for their own benefit. Black Studies is an interdisciplinary field of study par excellence, and it is well poised to bridge the chasm between the humanities, social sciences, and physical sciences.

Footnotes

Acknowledgements

The Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) analysis that provided the geochemical data for this article was completed by the second author at the Field Museum, Chicago. Both authors are grateful to Dr. Chapurukha Kusimba and Dr. Laure Dussubieux of the Field Museum for their valuable cooperation. The first author expresses his gratitude to the National Commission for Museums and Monuments (NCMM) for providing the permit to conduct archaeological research in Osun Grove. His special gratitude goes to the following NCMM officers: Mr. Oluremi Adedayo (Director of Heritage Sites and Monuments), Dr. Seyi Hambolu (Director of Research, Planning and Publications), and Mr. Olakunle Makinde (Site Manager and Curator, Osun Grove). For access to additional instrumental laboratories, the authors acknowledge the generous assistance of Dr. Craig Allan and Dr. Andy R. Bobyarchick of the Department of Geography and Earth Sciences at UNC Charlotte. An earlier version of this article was presented at the 2015 National Council for Black Studies conference in Los Angeles. The authors are grateful to those who attended the session or their helpful feedback.

Declaration of Conflicting Interests

The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: All the interpretations offered in this article are conclusions reached by the authors alone. The funding agencies and research laboratories have not in any way influenced their conclusions.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The first author gratefully acknowledges that the archaeological fieldwork that formed the backbone of the research was funded by a Dumbarton Oaks (Gardens and Landscape) summer research grant (2011), a Wenner-Gren Foundation grant (7099), and a National Endowment for the Humanities faculty research grant (HR-50114-04). The Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) analysis was made possible by the Field Museum’s Robert O. Bass Fellowship (awarded to the second author) and by the University of North Carolina at Charlotte (Office of the Dean of the College of Liberal Arts and Sciences) research grant (awarded to the first author). The latter paid for the LA-ICP-MS analysis of the samples reported here.

1.

Omotoso Eluyemi was a professor at the Obafemi Awolowo University, Ile-Ife (Nigeria), 1973-1993, where he taught archaeology and led several archaeological and oral historical investigations on the cultural history and technology of ancient Ile-Ife. He was the founding chair of the Department of Archaeology in the university in 1984. The first author was Dr. Eluyemi’s student between 1984 and 1988 during which he took several of the professor’s lectures and served on his research projects including the one that focused on the ethnography and archaeology of glass manufacture in Ile-Ife.

Author Biographies

Akinwumi Ogundiran is chair of the Africana Studies Department, and professor of Africana Studies, Anthropology & History at the University of North Carolina, Charlotte. His primary scholarship focuses on the archaeology, cultural history, and historical anthropology of the Yoruba world, Atlantic Africa, and the African Diaspora.

O. Akinlolu Ige is a professor in the Department of Natural Sciences, Middlesex County College, Edison, NJ. He is a former director of the Natural History Museum at Obafemi Awolowo University, Ile-Ife (Nigeria) where he also served as Professor of Geochemistry, Material Conservation and Museum Studies.

References

Ajayi

T. R.

Suh

C. E.

(1999). Partially extractable metals in the amphibolites of Ife-Ilesha area: A lithogeochemical approach for gold extraction. Journal of Mining and Geology, 35, 103-116.

Drewal

H. J.

Mason

(1998). Beads, body, and soul: Art and light in the Yoruba universe. Los Angeles: UCLA Fowler Museum of Cultural History.

Dussubieux

Gratuze

(2003). Non-destructive characterization of glass beads: An application to the study of glass trade between India and Southeast Asia. In Karlstrom

Kallen

(Eds.), Fishbones and glistering emblems: Southeast Asian archaeology (pp. 135-148). Stockholm, Sweden: Museum of Far Eastern Antiquity.

Eluyemi

(1987). The technology of the Ife glass beads: Evidence from the Igbo-Olokun. Odu, 32, 200-216.

Emofurieta

W. O.

Aladesewa

A. I.

Ogunseji

(1995). Secondary geochemical dispersion and mineralogical dispersion pattern associated with lateritization process in Ile-Ife, Southwestern Nigeria. Journal of Mining and Geology, 31, 39-52.

Freestone

I. C.

(2006). An indigenous technology? A commentary on Lankton et al. “Early primary glass production in southern Nigeria.” Journal of African Archaeology, 4, 139-141.

Ige

O. A.

(2009). Ancient glass making in Ile-Ife, Southern Nigeria. In Ignatiadou

Anastassios

(Eds.), Annales du 18e Congrès de l’Association Internationale pour l’Histoire du Verre (pp. 486-490). Thessaloniki, Greece.

Ige

O. A.

(2010, October, 3-6). Characterization and preservation of glass beads from Ile-Ife, southwestern Nigeria. In Roemich

(Ed.), Glass and ceramic conservation 2010. Proceedings of Interim Meeting of the ICOM-CC Working Group (pp. 66-72). New York, NY: ICOM Committee for Conservation in association with the Corning Museum of Glass.

Ige

O. A.

Durotoye

Oluyemi

E. A.

(2005). Mineralogy and geochemistry of lateritic profiles on ultramafic rock bodies around Mokuro in Ile-Ife, southwestern Nigeria. Journal of Mining and Geology, 4, 11-18.

10.

Ige

O. A.

Rehren

(2003). Black sand and ironstone: Iron smelting in Modakeke, Ife, Southwestern Nigeria. Journal of the Institute of Archaeo-Metallurgical Studies, 23, 15-20.

11.

Lankton

W. J.

Ige

O. A.

Rehren

Th.

(2006). Early primary glass production in southern Nigeria. Journal of African Archaeology, 4, 111-138.

12.

Ogundiran

(2002). Of small things remembered: Beads, cowries, and cultural translations of the Atlantic experience in Yorubaland. International Journal of African Historical Studies, 35, 427-457.

13.

Ogundiran

(2014). The making of an internal frontier settlement: Archaeology and historical process in Osun Grove (Nigeria), 17th–18th centuries. African Archaeological Review, 31, 1-24.

14.

Rehren

Th.

Freestone

I. C.

(2015). Ancient glass: From kaleidoscope to crystal ball. Journal of Archaeological Science, 56, 233-241.

15.

Willett

(1967). Ife in the history of West African sculpture. London, England: Thames & Hudson.

16.

Willett

(1977). Baubles, bangles and beads: Trade contacts of medieval Life: Thirteenth Melville J. Herskovits memorial lecture. Edinburgh, Scotland: Edinburgh University Press.