著者

遠部 慎 宮田 佳樹 小林 謙一 松崎 浩之 田嶋 正憲

出版者

国立歴史民俗博物館

雑誌

国立歴史民俗博物館研究報告 (ISSN:02867400)

巻号頁・発行日

vol.137, pp.339-364, 2007-03

---

岡山県岡山市(旧灘崎町)に所在する彦崎貝塚は,縄文時代早期から晩期まで各時期にわたる遺物が出土している。特に遺跡の西側に位置する9トレンチ,東側に位置する14トレンチは調査当初から重層的に遺物が出土し,重要な地点として注目を集めていた。彦崎貝塚では土器に付着した炭化物が極めて少ないが,多量の炭化材が発掘調査で回収されていた。そこで,炭化材を中心とする年代測定を実施し,炭化材と各層の遺物との対応関係を検討した。層の堆積過程については概ね整合的な結果を得たが,大きく年代値がはずれた試料が存在した。それらについての詳細な分析を行い,基礎情報の整理を行った。特に,異常値を示した試料については,再測定や樹種などの同定を行った。結果,異常値を示した試料の多くは,サンプリング時に問題がある場合が多いことが明らかになった。特に水洗サンプルに顕著で,混入の主な原因物質は現代のものと,上層の両者が考えられる。また,混入した微細なサンプルについても,樹種同定の結果,選別が可能と考えられた。これらの検討の結果,明らかな混入サンプルは,追試実験と,考古学的層位などから,除くことが出来た。また,9トレンチと14トレンチと2つのトレンチでは堆積速度に極端な差が存在するものの,相対的な層の推移は概ね彦崎Z1式層→彦崎Z2式層→中期層→彦崎K2式層→晩期ハイガイ層となることがわかった。今後,本遺跡でみられたコンタミネーションの出現率などに留意しつつ,年代測定試料を選別していく必要がある。そういった意味で本遺跡の事例は,サンプリングを考えるうえでの重要なモデルケースとなろう。Relics from all of the periods from Earliest Jomon through to Latest Jomon have been excavated from the Hikosaki shell midden situated in Okayama City (formerly Nadasaki-cho) in Okayama Prefecture. Multiple layers of relics were excavated at the start of the survey from trench 9 located on the western side of the site and trench 14 on the eastern side of the site, which drew attention to these locations as important spots. Although extremely few pieces of carbonized material adhering to pottery were found in the Hikosaki shell midden, large quantities of carbonized wood were recovered from excavation. We undertook dating mainly of the carbonized wood and investigated the corresponding relationships between the pieces of carbonized wood and relics from each of the layers. Although the results were largely consistent in terms of the sedimentation process of the layers, there were some samples whose dates deviated considerably. We made a detailed study of these samples and sorted basic information. We repeated measurements and identified the tree types of those samples that gave abnormal readings.The results revealed that for many of the samples that yielded abnormal values there were problems at the time of sampling. This was particularly pronounced for samples that were washed in water. It is conceivable that the substances that caused this contamination were both present-day substances and the top layer. Identification of the types of wood of very small contaminated samples also showed that screening is a possible cause of this contamination. As a result of such investigations, we were able to eliminate samples that were clearly contaminated from additional testing and the archaeological succession of strata. Although there was an extreme difference in the sedimentation rates for trench 9 and trench 14, we found that the pattern of relative layer development was mainly : Hikosaki Z 1-type layer -> Hikosaki Z 2-type layer -> Middle Jomon layer -> Hikosaki K 2-type layer -> Latest Jomon Haigai layer.In the future, it will be necessary to screen samples for dating while keeping in mind the contamination rates seen at this site. In this sense, the case of the Hikosaki site will become an important model with respect to sampling.