UDC 947:940.1 +630.561.1.24

The article is devoted to the dating of a unique monument of wooden architecture of the era of Russian colonization of Siberia - the Saviour's Church from the polar city of Zashiversk. For the first time, dendrochronological analysis was used to study the church. This method, based on the variability of the growth of annual tree rings, allows you to get the most objective dates with an accuracy of up to a year. A record-breaking sample of 95 logs from different walls and crowns in each room of the church was studied. To establish calendar dates, we used generalized tree-ring chronologies obtained from live larch trees in the area adjacent to the monument. Based on the established dates of felling trees that have preserved the podkorov ring, we can say that the bulk of the wood was harvested in the period 1709-1711. The construction of the building itself could have taken place either at the same time or immediately after 1711. It cannot be ruled out that some logs with earlier dates were first part of the previous structure and were later reused for the construction of the real Saviour's Church from Zashiversk.

Keywords: monuments of wooden architecture, Siberia, history, architecture, dendrochronology.

Introduction

Among the few surviving monuments of wooden architecture from the era of Russian colonization of Siberia, the Saviour's Church from Zashiversk occupies an outstanding place (Figs. 1, 2). It was transported in 1971-1972 from the banks of the river Zashiversk. 3) to Novosibirsk and later restored here, it became the central exhibit of the Historical and Architectural Museum of the Institute of Archeology and Ethnography of the Siberian Branch of the Russian Academy of Sciences. Summarizing the results of archaeological and historical research, the characteristics of the architectural and construction features of the unique tent church are described in detail in the capital work of A. P. Okladnikov, Z. V. Gogolev and E. A. Ashchepkov [1977]. However, many questions related to the time of construction of the church and its subsequent history remained unclear. This aspect is most thoroughly discussed in a number of recently published works, where the authors proposed their own versions of the answers based on the analysis of a complex of various data: written and graphic sources, the results of archaeological excavations and a full-scale survey of the building (Kurilov and Mainicheva, 2005, p.29).

At the same time, it is striking that so far, among the methods and approaches used, the most obvious one (since we are talking about a building made of wood) has not been used - dendrochronological analysis, which is based on recording the variability of annual wood growth and allows us to date events with high time resolution (year, season) [Shiyatov et al., 2000, p. 15]. As applied to wooden architectural monuments, this method helps to solve a wide range of issues most accurately and reliably: from confirming or clarifying the date of their construction to restoring a complete picture of the construction history over long periods of time, as well as reconstructing numerous aspects of woodworking and construction [Chernykh, Sergeeva, 1997, p.109].

The Institute of Archaeology of the Russian Academy of Sciences has gained extensive experience in using dendrochronological analysis in studying Russian architectural monuments of the 15th and 20th centuries in the northern regions of European Russia. In labo-

The research was carried out with the financial support of the Russian Foundation for Basic Research (project N 08 - 06 - 00429) and RGNF (project N08-01-18094).

1. Saviour's Church from Zashiversk with a bell tower, View from the northeast. Photo by V. P. Mylnikov.

2. Saviour's Church from Zashiversk. View from the southeast. Photo by D. A. Mainichev.

3. Location of Zishiversk and districts for which generalized tree-ring chronologies were obtained (UJD, ZHAS, MOMA).

A special program of such research has been developed in the laboratory of natural science methods, which currently covers hundreds of buildings [Chernykh and Sergeeva, 1997; Chernykh, 1996, 2001; Chernykh and Karpukhin, 2006]. As noted by N. B. Chernykh, who has been promoting this work for many years, "a comprehensive study of wooden architectural monuments (including the restoration of religious buildings) in recent years can no longer be imagined without the use of dendrochronology methods" [Chernykh and Sergeeva, 1997, p.109].

Some Siberian architectural monuments also served as objects of dendrochronological research. This is primarily the Kazym prison, where a pioneering work on dating two towers and other structures was carried out (Komin, 1980). In recent years, experts from the Laboratory of Dendrochronology of the Institute of Plant and Animal Ecology of the Ural Branch of the Russian Academy of Sciences in Yekaterinburg have been active in the field of dating architectural monuments of the Urals. For example, they obtained dates for religious and civil buildings of the 18th century on the territory of Verkhoturye from annual rings (Goryachev, 1997, 1998).

Thus, the need for a detailed dendrochronological study of the Saviour's Church from Zashiversk is long overdue.

Materials and methods

The dendrochronological analysis was based on wood samples from the logs of the Saviour's Church from Zashiversk, selected in 2006 in the Open-air Museum of the Institute of Electrical Engineering of the Siberian Branch of the Russian Academy of Sciences by researchers B. C. Myglan (Siberian Federal University) and I. Y. Slyusarenko( Institute of Electrical Engineering of the Siberian Branch of the Russian Academy of Sciences), as well as sawn wood taken by N. D. Ovodov in 1971 from the northern side the temple*.

* The authors are grateful to A. P. Borodovsky for providing information on the location of spil.

Table 1. Results of cross-dating of Spasskaya Church wood samples from Zashiversk

Continuation of Table 1

End of Table 1

Notes: t is the sensitivity coefficient; σ is the standard deviation; the following abbreviations were used to designate logs: S, Z, Y, V are the initial letters of the cardinal directions, Bp are samples taken from the roof level, P is the partition between the refectory and the quad; the logs of the octagon are marked according to metal plaques on them.

Considerable difficulty in sampling was caused by the incredible hardness of the wood (larch), which in some cases did not allow taking samples from the intended areas of logs, leading to the destruction of not only the peripheral rings, but also the core itself. Despite this, as a result of the work carried out, it was possible to select material from 95 logs for desk processing, of which two were later not taken into account. Thus, our sample is a record in terms of representativeness, if we take into account that in normal practice, the number of samples from one object rarely exceeds 25-30 pcs. [Chernykh, 2001, p. 115]. In order to make the analysis as detailed as possible, the material for the study was selected from different walls and crowns in each room of the church. According to the architectural elements of the building, the samples were distributed as follows (Table 1): chetverik-22 from the north wall, 2 from the south; refectory-9 from the north wall, 20 from the west, 8 from the south; partition between the chetverik and refectory-9; octagon-21; apse-1 (location most of the samples are shown in Figs. 4, 5).

In order to obtain accurate tree felling dates ("felling dates"), we paid close attention to the presence of the last ring formed during the life of the tree when selecting samples. Therefore, sections of logs were selected without visible signs of surface damage or with obvious signs of an outer ring (bark remnants, etc.). As the study showed, such rings were best preserved in logs inside the room, especially in the attic, unvisited part of the building, since here they were most protected from atmospheric and other influences.

Under laboratory conditions, the surface of the samples was cleaned with a scalpel. The width of the annual rings was measured from the center to the periphery on a semi-automatic LINTAB-2 unit with an accuracy of 0.01 mm. In order to establish calendar (absolute) dates for individual samples and the structure as a whole, we used generalized tree-ring chronologies obtained from live larch trees: ZHAS (LarixsibiricaLdb, 1591-1991, 67°27 'N, 142°37' E), UJD (Larix sibirica Ldb, 1369-1991, 68°41 'N, 143°16' E), MOMA (Larix sibirica Ldb, 1311-1945, 65°53 'N, 145°18' E)*. These chronologies characterize the growth of trees-

* Taken from the database of the laboratory "Structure of annual rings" of the Forest Institute SB RAS and [Schweingruber, 2006].

Fig. 4. Fragment of the facade (without gallery) on the north side. Numbering and dates of samples according to Table 1.

Fig. 5. Fragment of the facade (without gallery) on the west side Numbering and dates of samples according to Table 1.

3). Cross-correlation analysis was performed in the specialized dendrochronological research package DPL (Holmes, 1984), and graphical cross - dating was performed in the TSAP V3 software package.5 [Rinn, 1996] with a visual comparison of the variability curves of absolute and indexed values of radial increment, calculations and graphic illustrations - in the program "Statistica 6.0".

Results and discussion

Using a combination of cross-correlation analysis and graphical cross-dating, the measured individual growth series were dated relative to each other and generalized tree-ring chronologies ZHAS, UJD, MOMA (Table 1).

Indicators such as the sensitivity coefficient (Ferguson, 1969) and the standard deviation were used to estimate the variability of individual series. The first characterizes the relative value of the weather variability of growth, the second-its amplitude. Quality control of dating was performed in the COFECHA program. High values of the sensitivity coefficient (average 0.40) and interseries correlation coefficient (average 0.67) indicate the presence of a clear climatic signal in changes in the width of annual rings and indicate a high degree of consistency in the growth of individual series. Binding of the latter to the absolute time scale was achieved by cross-dating them relative to the generalized tree-ring chronologies UJD and MOMA (Fig. 6). As a result of this work, calendar years corresponding to the outer annual rings preserved on the samples were determined. However, the last peripheral rings do not always reflect the years of felling of trees ("felling dates") used in the construction of the church. In the vast majority of samples, after the last fixed ring, an unknown number of outer rings were lost for one reason or another.

Of the 93 logs examined, only 14 (15% of the total) have a reliable crustal layer, of which 13 are from the octagon and one from the northern wall (cut 1971, taken during the analysis of the church). Of the 14 trees with the established year of felling, 10 were felled between 1709 and 1711. In total, these years account for 17 samples (18%). At the same time, when estimating the distribution of peripheral rings over decades (Figure 7), we note that in 1700-1711 36.5 % (34 samples), and in 1690-1699. -

Figure 6. Example of cross-dating of individual series relative to the generalized MOMA chronology.

1-sample 79; 2 - 80; 3 - 32; 4 - 58; 5 - 93; 6 - MOMA.

Sample numbers according to Table 1.

Figure 7. Histogram of the distribution of peripheral rings.

25 % (23 samples). In other words, more than 60% of dates fall within the last 20-year interval.

An analysis of the distribution of dates by church premises showed the following. For the apse there is a single date - 1710; for the chetverik the latest-1701; for the refectory-1709; for the partition between the chetverik and the refectory-1706, for the octagon-1711. The greatest number of the latest dates of tree felling (1709-1711), as indicated above, is associated with the octagon. At the same time, the presence of logs with the "felling dates" of 1685, 1702, and 1706 (samples 73, 76, and 88) indicates that the construction of the church used either dead trees or logs harvested in earlier years or used again.

As already noted, most of the analyzed samples have lost their last life rings. This is due to a number of reasons, including weathering of the surface of logs in the harsh conditions of Indigirka, their possible rafting on the river, disassembly and transportation to Novosibirsk, storage for a number of years, and new assembly on the territory of the Open-air Museum. To find out how many rings could have been lost as a result of various external influences, a series of consecutive measurements were carried out for several radii of the cut taken in 1971 (sample 32) and preserving the last lifetime ring (fragments of the crust were present in one area). The results obtained showed that the difference between the radius containing the subcrustal ring (1709) and the last peripheral rings of other radii (without obvious signs of damage on the outer side) can reach 10-18 years. Thus, at least for some of the logs, their earlier dates, which fall in the last decades of the XVII -first years of the XVIII century, should be corrected taking into account the revealed difference, then the number of dates close to the period 1709 - 1711 will increase. It can be argued that during this period, most of the timber was harvested for the Saviour's Church from Zashiversk, and the construction itself could have begun either in the same years or immediately after 1711.

There is an assumption that the Zashiversky prison was located at the mouth of the Moma River before 1700, and a small chapel played the role of a religious building. After 1700, the main structures of the old prison, including the chapel, were moved downstream of the Indigirka River by rafting. At the new site, the chapel was first significantly rebuilt, and then became part of the building of the Saviour's Church [Kurilov and Mainicheva, 2005, p.40-43]. It was of interest to test this assumption by analyzing the extent to which the increase in samples from Tserkva (respectively, the area of wood origin) is consistent with the generalized tree-ring chronologies obtained for sites close to Zashiversk and located along the north-south line, including in the lower reaches of the Moma River (see Fig. 3). A comparison of the growth indices showed that the greatest synchronicity of the curves (due to the impact of general external conditions) is observed with the MOMA chronology (Table 2). This allows us to conclude that the site of harvesting trees used for the construction of the church was located higher along the Indigirka from Zashiversk. This fact has two possible explanations.

1. Initially, the prison could actually be located at the mouth of the Moma River, and then it was dismantled and floated to the Indigirka River. In this case, the explanation is:-

Table 2. Values of the paired correlation coefficient between the indexed tree-ring chronology of the Zashiver Church and indexed generalized chronologies

Note: N is the amount of sample used (number of years); p < 0.05, all values of correlation coefficients are significant.

It is assumed that there are groups of earlier dates of church logs in the interval 1660-1690s. Although the outer rings of these logs are missing, it can be assumed that their dating to the last decades of the XVII century is associated not so much with the loss of rings, but with the harvesting of wood during this period and its use in the construction that preceded the church itself.

2. Wood for the construction of the temple was harvested much upstream, and then floated down the river. This possibility is indirectly indicated by the impressions of the participants of the expeditions of 1969-1971: looking at the powerful larch logs from which the church was cut down, they noted that "such trees do not grow in the forest tundra near Zashiversk. They must have been fused from far away, from the upper reaches of the Indigirka River" [Okladnikov, Gogolev, and Ashchepkov, 1977, p. 39]. The lack of suitable construction wood on the site of the modern location of Zashiversk is also indicated by the evidence of travelers of the XVII century: "From Zashiversk... the forests are thinning and the whole country is one vast lake... Here there is nothing but moss, the invariable gnarled larch, even smaller talus and water" (see: [Ibid., pp. 48-49]).

Conclusion

Based on the dates set for the Spasskaya Church for cutting down trees that preserved the podkorov ring, we can say that the bulk of the wood was harvested in the period 1709-1711. The building itself could have been constructed either at the same time or immediately after 1711. In any case, no date later than 1711 was recorded in the entire very large sample. In our opinion, this indicates that the existing timbers were used for possible alterations for the chetverik, octagon and refectory of the church. It cannot be ruled out that some of the logs with earlier dates were first part of the previous construction and were later used for the construction of the real Saviour's Church from Zashiversk.

List of literature

Goryachev V. M. Dating of wood samples from archaeological excavations on the territory of the Nikolayevsky monastery in Verkhoturya // Protected archaeological research in the Middle Urals. Yekaterinburg: Yekaterinburg, 1997, issue 1, pp. 166-174.

Goryachev V. M. Nekotorye itogi datirovki ostankov derevennykh stroeniy iz arkheologicheskikh raskopov na territorii g.Verkhoturya [Some results of dating the remains of wooden structures from archaeological excavations in the territory of Verkhoturya]. Yekaterinburg: Bank of Cultural Information, 1998, pp. 6-13.

Komin G. E. Dendrochronology of Kazymsky gorodok / / Historical and architectural museum under the open sky: principles and methods of organization. Novosibirsk: Nauka Publ., 1980, pp. 121-126.

Fortress and Temple: Trends in the movement of Russian culture in architectural creativity. Novosibirsk: Izd-vo IAET SB RAS, 2005, 96 p. (in Russian)

Okladnikov A. P., Gogolev Z. V., Ashchepkov E. A. Drevny Zashiversk: drevnerussky zapolyarny gorod, Moscow: Nauka Publ., 1977, 212 p.

Chernykh N. B. Dendrochronology and archeology, Moscow: Nox, 1996, 216 p.

Chernykh, N. B., Dendrochronology of architectural monuments of Karelia, KSIA, 2001, issue 211, pp. 113-120.

Chernykh N. B., Karpukhin A. A. Construction of the" old town " of the Kirillo-Belozersky monastery according to dendroanalysis data (Ivanovsky Monastery) / / RA. - 2006. - N1. - pp. 157-163.

Chernykh N. B., Sergeeva N. F. Dendrochronology of architectural monuments in the north of Russia (basin of the Northern Dvina and Onega rivers) / / RA. - 1997. - N 1. - pp. 109-124.

Shiyatov S. G., Vaganov E. A., Kirdyanov A.V., Kruglov V. B., Mazepa V. S., Naurzbaev M. M., Khantemirov R. M. Metody dendrochronologii: ucheb.-method. stipend. Krasnoyarsk: Krasnoyar. State University, 2000. - Ch. 1: Fundamentals of dendrochronology. Collecting and receiving tree-ring information. - 80 s.

Ferguson C.W. A 7104-year annual tree-ring chronology for Bristlecone pine, Pinus aristata, from the White Mountains, California // Tree-Ring Bull. - 1969. - Vol. 29, N 3/4. - P. 3 - 29.

Holms R.L. Dendrochronological Program Library/ Laboratory of Tree-ring Research. - Tucson: The University of Arizona, 1984. - 51 p.

Rinn F. TSAP V3.5: Computer program for tree-ring analysis and presentation. -Heidelberg: Frank Rinn Distribution, 1996. -269 p.

Schweingruber E. Zhaschiviersk, Moma River. - 2006. -URL: http://www.ncdc.noaa.gov/paleo/treering.html

The article was submitted to the Editorial Board on 09.06.08.

Permanent link to this publication: