Phase II and Phase III Archaeological Database and Inventory
18BC56 The American Can Company
Diagnostic Artifact Data
Features
Summary
The American Can Company (18BC56) consists of the archeological remains associated with a 19th-20th century industrial complex in the Canton area of Baltimore, Maryland. The site encompasses 9 1/3 acres. It is a triangular lot bounded by Boston Street to the south, Lakewood Avenue to the east, and Hudson Street to the north. By the time of the first archeological investigations in the 1990s, the complex was inoperative, but several buildings and warehouses remained. The core of the complex consists of a group of structures which cross-cut the site center, north-south. Of this group, the earliest construction is known as the Norton Building (ca. 1895). This structure adjoins the Manufacturing Building (ca. 1913) to its north and the 1902 Warehouse to its south. An additional structure at the intersection of Boston and Hudson Streets known as the Machine Services Building or “Signature Building”, a 1924 construction, is a familiar Baltimore landmark. A1952 Shipping Warehouse in the southern section of the site had been mostly demolished. Most of the area surrounding the site is heavily developed today. Soils mapped for the area are classified as “Urban Land” meaning that 80% of the surface is covered either by buildings or by impervious surfaces such as asphalt or concrete.
A 1935 map of Baltimore depicts the original shorelines and stream courses within the city. According to this map Harris Creek encompassed the entire American Can Company (hereafter ACC) property to north of Hudson Street and east of Lakewood Avenue. The location of the site was inundated for much of the 19th century.
The major development of the Canton area came at the behest of the Canton Company of Baltimore, which formed in 1829 when a group of businessmen purchased 2,000 acres east of Fell’s Point as a real estate venture. The result was an early precursor of modern-day industrial park development in Canton. The Canton Company devoted significant energies and money into re-routing natural streams, infilling areas to create new land, lumbering, and sub-dividing lots to develop on their original investment. These lots were then leased or sold to recoup costs and earn a profit. Changes to the natural drainages of the area also came at the insistence of the City of Baltimore, low-lying areas of which were frequently flooded and needed better storm-water management.
The sequence of in-filling Harris Creek is not clear from historic maps, but some patterns can be distinguished across time. All maps show the Creek narrowing at Hudson Street. At this point a bridge was constructed by 1829. According to City Ordinances, as early as 1821 portions of Harris Creek were drained and filled above Wilkes Street. City Commissioner’s Annual Reports between 1878 and 1887 also indicated street construction and Creek in-filling prior to this time. By 1880, a Sanborn Fire Insurance Map shows a diversion of the Creek westward. The site area was filled-in by then and occupied by George Tyler’s sawmill. In 1887, the City Engineer remarked in his annual report that the volume of water in Harris Creek posed unique problems for drain construction at the time. Similarly, the 1888 report indicated that the Creek still held water, and it was necessary to drive piles and operate pumps to complete construction of a city drain to Harford Run.
In short, Harris Creek encompassed the site area until at least 1850, but by 1880 the portion east of Luzerne was in-filled. By 1896, the entire Creek had been in-filled or diverted into the Harford Run Storm Water Drain, completed in 1888.
The materials used for in-fill practices of the late 19th century are documented by various sources. Canton Company records reported the use of oyster shell. The City Engineer in 1888 explicitly referenced oyster shell, as well as “elegant filling” from cellar excavations of nearby residences, and refuse from Abbott’s Iron Rolling Mill. Other sources reveal that materials dredged from the harbor, coal, ashes, shell, and soils were also used. This created the land on which the American Can Company and other Canton area industries were built.
Baltimore’s canning and can-manufacturing industries became world renowned during this period and Boston Street in Canton was known as “Canners Row”. Until the mid-1800s, can manufacture was subsumed under the canning, or food-processing industry. Tin Cans were created manually by tinsmiths, filled, and sealed. By 1888, in response to canning industry growth, can manufacture had emerged as a separate industry. Norton Tin Plate and Can Company, a Chicago-based manufacturer, established its operations in Canton (at the site) in 1888. In 1883 Norton had introduced mechanization to the industry through their invention of a semi-automatic machine for soldering can seams, and the industry had become fully automated with the introduction of the “sanitary can” in 1898. Norton merged with other can manufacturers in 1901 to form the American Can Company. At the time of the merger, Norton was considered the largest can manufacturer in the country. They would continue to operate out of Canton until the mid-20th century.
The only documented archeological work to occur at 18BC56 was in 1993 and 1994. The investigation was undertaken to assess cultural resources in the vicinity of the Harford Run Storm Water Drain where it crosses the American Can Company property in the vicinity of the former shipping warehouse (see above). This area would be impacted by the (then) proposed construction of the Lakewood Avenue Storm Drain extension. Investigations were conducted in compliance with Section 106 of the National Historic Preservation Act of 1966, as amended.
Two areas were selected for excavation. Area A consisted of a 15.24 m (50 ft) by 6.1 m (20 ft) asphalt covered section at the base of the Shipping Warehouse, and 6.1 m east of the imprint of the Harford Run Drain. Area B consisted of a 4.6 m (15 ft) square, 12.2 m (40 ft) east of Area A, within the concrete footprint of the warehouse. These areas were mapped, and elevations were recorded with a transit. Both areas were then mechanically excavated with a backhoe with a 76.2 cm (2.5 ft) wide bucket. In Area A, four trenches were excavated. One trench was excavated on an east-west axis. The remaining trenches were excavated on a north-south axis. This arrangement permitted a wide cross-section view of the soil stratigraphy. In Area B, the entire 4.6 m square was excavated.
Systematic excavation and sampling methods were implemented to better assess functional and temporal aspects of the fill sequence. Excavation proceeded as follows: soil strata were individually removed and segregated into piles. Three sampling strategies were employed: A standardized sample (91 cm square and 15 cm deep) of a single stratum from each trench was collected and screened through hardware mesh. Shell, brick, slag, cinder, and coal were counted and weighed and discarded at the site. The remaining artifacts were collected. This type of sampling enabled researchers to determine whether a particular fill layer was functionally and/or temporally intact. It could potentially provide indications of the fill sequence.
After excavation was completed, column samples were collected from the west walls of each trench. These were excavated by 15.24 cm (6 in) square sizes. Column samples can provide information about artifact densities within each stratum and, like screened samples, provide indications of the land-fill sequence. They may also provide useful information about the integrity of particular fill layers. These samples were screened and all artifacts were collected. Finally, a five-gallon bucket-size judgment sample of artifacts was collected from each trench. Judgment sampling is a subjective method of artifact collection which provided an estimate of the kinds of materials present, especially large, highly visible artifacts.
Soil profiles were drawn and photographed. Soils were recorded according to color and consistency. Intact strata or in situ features were manually excavated with shovels and tamped. Features were photographed and mapped into the site plan. Subsequent to excavation, all trenches were re-filled and tamped. The exposed soil was resurfaced with tar and chip.
Excavation of the four trenches in Area A and a large block in Area B demonstrated that fill soils constituted the subsurface of the area to be impacted by the drain extension. These soils were likely brought to the area for purposes of creating land at the location of Harris Creek in the late 19th century. The deepest levels of fill (an oyster shell/silt mix) and the degree of slope were strongly indicative of the old shorelines of Harris Creek. It appeared that after the Creek was filled, and perhaps concurrent with the construction of the Dillon Street lateral drain ca. 1885, artificial land was created. The artificial land was subsequently disturbed during the construction of the ACC shipping warehouse around 1952.
Based on observations of stratigraphy and artifact content, the following sequence is proposed. An oyster shell fill was laid in the creek bed ca. 1880, or prior to drain construction. The Drain was built, then sealed in clay and covered with several additional layers of fill. During the 1952 construction of the ACC shipping warehouse, deep piles were driven into the fill and the building footprint was constructed, thus causing major soil disruption. This was clear from the sequence in the large block excavation in Area B where builder’s sands were intermittently mixed with an otherwise organic fill. Soils in both areas were also disturbed at this time by the installation of two water pipes (possibly others). At some unknown time, a footer/support was added to the base of the Dillon Street Drain, causing further soil disturbances. It was proposed that the most severe disturbances occurred in the 20th century when heavy machinery literally inverted portions of the fill strata. Twentieth century activities may explain the presence of modern bottle glass and Styrofoam in otherwise consistent 19th century levels.
In short, the soils observed in excavation demonstrated late 19th century fills disturbed by 20th century activities. Additional 20th century fills were added when the property was graded and surfaced.
The Late 19th century fills are exemplified by a layer just above the level of the drain installation containing a mixture of industrial and domestic refuse. A disproportionate amount of slag strongly suggests a furnace source, possibly Abbott’s Iron Works. City Commissioner’s reports refer to Abbott’s as a source for fills. Domestic refuse obtained from local cellars is also referenced in these reports, although the point of origin could not be substantiated. What is difficult to explain is the nature of the slag-domestic refuse mixture. The level did not exhibit night soil characteristics (brown/black organic qualities) expected of a domestic context, yet artifact size, content (bottles, ceramics, bone), and degree of preservation suggest a midden-like origin. Prior to in-filling, the slag and domestic refuse may have been mixed. There were no visible indicators that this occurred after deposition.
Historic reports also document late 19th century in-filling procedures which involved hauling refuse in wagons. This might explain a high frequency of intact vessels and large mendable ceramics. In a situation where land is created by manual labor, one would expect less breakage, in comparison with that expected from the use of heavy machinery.
A total of 30,934 artifacts were collected at the site. Given that the context of the vast majority of these finds is in fill soils derived off-site, the artifacts have been cataloged in the table above simply as miscellaneous objects rather than by their functional categories. Within the ceramic assemblage were 6 creamware sherds, 32 pearlware sherds, 29 porcelain sherds, 45 redware sherds, 27 yelloware, 138 whiteware, 441 ironstone, 66 stoneware, and 68 unidentified sherds.
Based on the findings, no additional archeological work was recommended at 18BC56. Based on the ubiquitous presence of secondary deposits from 19th and 20th century fills, the site likely has very little research potential.
|
