Patrick M. Malone

Surplus Water, Hybrid Power Systems, and Industrial Expansion in Lowell

Lowell is famous as the first great industrial city built at an American waterpower site. The figures usually given for the amount of waterpower provided by the Pawtucket Dam and the Lowell Canal System are based on the documented leases of "permanent" millpowers. This was waterpower available all year, even during dry seasons when flow in the Merrimack River was reduced. In reality, Lowell's corporations regularly made use of much more flow than that minimum contractual obligation. "Surplus water" helps to explain the continuing increase in production in Lowell after the Civil War. Its existence was a significant stimulus for the installation of boilers and steam engines in the mill complexes on the canal system. Both material and documentary evidence show that steam power and waterpower worked together to drive the growing number of spindles in the city. Without surplus waterpower, Lowell's industrial expansion might have stalled in the 1870s.

Preface

Material evidence should play an important role in almost any historical research project focusing on industrial development. Artifacts are, of course, the focus of investigations by industrial archaeologists, who by definition are concerned with the physical remains of the past and the information that "things" can provide. Most people know that archaeologists use objects or structures to answer questions about the creation of an industrial site or the work that was done there. Another critical contribution of material evidence is often overlooked: raising questions that require intensive archival research to answer. A sequence that seems common in the best examples of industrial archaeology is (1) a research question and/or a hypothesis formed in the presence of baffling physical evidence; (2) archival research using the documentary record, ideally done in association with repeated examination of the industrial site; (3) conclusions that are supported by both documentary evidence and "ground proofing." Few questions of historical significance can be answered by artifacts alone, but physical evidence or even an evocative "sense of place" can create the spark of intuition or curiosity that leads to an effective research project using a variety of evidence.

This study of surplus waterpower and its connections with steam power began in the mid-1970s in a strongly evocative but silent industrial space. I, with other members of a Historic American Engineering Record (HAER) recording team, had gained access to the abandoned wheel room of the former Suffolk Mills in Lowell, Massachusetts.¹ We saw at once that four stone wheel pits were still in place, their enormous cylindrical forms indicating that each had once held a single Boyden turbine. The pits dated from 1848, a year after the completion of the Northern Canal in Lowell. We were not surprised to see that more recent Francis-type (mixed-flow) turbines had replaced the outward-flow Boydens, but the number of wheels was unexpected. One pit had a single turbine, but in each of the other three pits there were a pair of turbines (see Figure 1). These high-capacity wheels, smaller but more powerful than the original Boydens, would have used a great deal of water, much more than the flow guaranteed by the Suffolk leases, which we had studied.² The physical evidence suggested that the amount of waterpower at the Suffolk Mills, and by extension on the canal system that the team was recording, might be considerably greater than historians had assumed.

Figure 1. Turbine pit in the wheel room of the Suffolk Mills. This stone pit is one of four in the room. Built in 1848 for one outward-flow Boyden turbine, it was converted to hold two mixed-flow Francis turbines that could take better advantage of surplus water. A metal penstock in the foreground delivered water to the vertical housing for both Francis units. The runner on the right has been removed, but gearing and shafting are still linked to the runner on the left. Photo by author.

Another 25 years of occasional research showed that surplus power (provided by extra flow above the leased quantities) was, indeed, a great asset for the mills on the Lowell Canal System. Additional insight came while I was showing some secondary school teachers³ the same Suffolk wheel room, which had become one of the interpreted sites in the Lowell National Historical Park. After I told the group that the seven installed turbines could handle not just the "permanent" flow in the Suffolk leases but also a substantial amount of surplus water, a teacher asked me why textile corporations would add extra production machinery that might sit idle for part of a year, if there was no surplus waterpower to run it in droughts. Thinking of the tall smokestacks that were still prominent on Lowell's skyline, I answered that steam engines would have supplemented or filled in for water turbines in droughts or floods (see Figure 2). However, I had never before recognized the need for interconnected prime movers in a place where surplus water was cheap but not always available. Then I went back to the archives to do more research on hybrid power systems in Lowell.

Figure 2. Brick chimney of the Boott Mills. This view, taken from a crane, shows the tall chimney added to the mill complex in the early 1880s. The brick structure reflects the increasing importance of steam power at this waterpower site. Courtesy of Charles Parrott.

Hydrology and Surplus Water

The hydrologic cycle was the basis for waterpower—an engine driven by a combination of gravity and solar energy. Water fell under the pull of gravity as precipitation into a river's drainage basin. Although some returned directly to the clouds through evaporation or transpiration, most of it proceeded downhill, either on the surface of the basin or through permeable soils and rock formations. Here again, the force of gravity played a critical role. Eventually most of the subterranean groundflow fed springs that added to the surface flow in streams or discharged directly into the sea.⁴ Elevation gave water potential energy as it moved down from interior uplands to tidal level. Water on the surface gave up that energy of position most dramatically at natural waterfalls and steep rapids or at the dams men placed in its path. When industrialists built their dams and canal systems to divert water through waterwheels, they were capturing part of the river's energy for production. The sun could always be counted on to replace the energy used, raising water vapor from lakes, streams, wetlands, and oceans and driving the cloud formations that fed the basin. The problem for manufacturers who wanted to operate mills all year was that the engine of the hydrologic cycle did not always run at a constant speed.

Not all rivers are suited for industrial development. Entrepreneurs in the 19th century sought mill sites with substantial head (drop) and relatively steady flow year round. In addition to developing factories and power canals, some industrialists worked to "improve" drainage basins by building dams and storing water in reservoirs for later release in dry seasons. Their principal goal was to increase the "permanent" (predictable minimum) waterpower that was available in all but an exceptional drought. Another goal, which most historians have ignored, was to make productive use of additional or "surplus" water. For much of the year, even a heavily engineered river carried a great deal more flow than the "permanent" quantity. Although most of that excess water went to "waste" (produced no power) by spilling over dams in freshets, it was possible to use a significant part of it for manufacturing.

The Merrimack River

On the Merrimack River, surplus water became a valuable commodity after the Civil War and proved to be a significant stimulus for both technological change and industrial expansion. However, for this unreliable source of extra waterpower to be useful in textile manufacturing, it had to be combined with steam power. Only the installation of steam boilers and engines could assure continued full production whenever the river had sharp reductions in surplus flow. In fact, only the promise of substantial amounts of surplus waterpower for most of the year made it worthwhile for textile corporations on the falls to add steam power and production machinery. Extra water could cut the consumption of fuel and provide major savings for industrialists in interior cities like Lowell, where coal was delivered by rail. It would have made little sense to build an entire steam-powered textile complex in Lowell, but adding a steam mill or a hybrid (combination of steam and water) mill to an existing water-powered complex could be a good business decision.

Lowell's great textile corporations had an effective strategy for powering their enlarged operations: they depended heavily on their leases of "permanent" waterpower, bought surplus waterpower whenever it was available, and used only as much steam power as they needed to maintain full production. This strategy explains how Lowell could continue to expand in the face of growing competition from a coastal city like New Bedford, which had no waterpower but received its steam coal by relatively inexpensive maritime transport.

The Merrimack flows down from the highlands of New Hampshire, past the great industrial cities of Manchester, Lowell, and Lawrence, and into the sea at Newburyport on the northern coast of Massachusetts. The watershed of the river extends northward to the flanks of Mount Washington. Near the summit of that peak are the headwaters of the Pemigewasset River. Dozens of streams and lakes feed that river as it runs south to Franklin, New Hampshire. There the Winnipesaukee River joins it from the northeast, bringing the full discharge of Lake Winnipesaukee and its southern bays. The junction of these two rivers forms the Merrimack, 269 vertical feet above the mean tide at Newburyport.⁵

Lowell: Building an Industrial City at the Falls

In its 110-mile journey to the sea, the Merrimack makes several dramatic drops. One of those sudden changes in elevation occurs at Pawtucket Falls in Lowell. For centuries, these falls produced a natural descent of almost 30 feet, "not perpendicular, but over several rapids, in circuitous channels, with a violent current amidst sharp-pointed rocks."⁶ Today, a masonry dam, topped by wooden flashboards, stands at the head of the falls. This structure diverts much of the water into power canals and adds to the natural drop through the rugged stone formations that stretch for more than one-half mile downstream.

In 1821, investors from Eastern Massachusetts were seeking a site with substantial waterpower potential, where they could build a new industrial city for the mass production of textiles. The success of the Boston Manufacturing Company, which some of them had helped to form in Waltham in 1813, gave them confidence as well as a desire for more space and waterpower. The place they selected for large-scale development was at Pawtucket Falls in a sparsely developed community known as East Chelmsford. After buying substantial tracts of land and a failing transportation canal that bypassed the falls, they began the construction of a power canal system, a machine shop, and the first of 10 textile corporations that would draw waterpower from the falls. These industrialists kept in place the corporation that had operated the old canal, the Proprietors of Locks and Canals (L&C), but they changed its role to serve the needs of the waterpower users on the system. By 1826, there were two levels of canals and 2,600 people in the newly incorporated town of Lowell, which was already one of the fastest growing communities in the United States. The manufacturing firms on the Lowell Canal System became the sole owners of the Proprietors of Locks and Canals in 1845 and soon began a series of acquisitions and engineering projects to assure better year-round flow in the Merrimack River.⁷

In combination with the Essex Company of Lawrence, representatives of the Lowell corporations gained control of more than 100 square miles of lake surface in New Hampshire in the years between 1845 and 1859. They stored water in the lakes during the usually wet winter and spring and called for releases whenever the Merrimack needed additional flow to meet the downstream demands of industry.⁸ With this vast reservoir system helping to even out the natural fluctuations in runoff, the Proprietors of Locks and Canals felt confident enough to lease more permanent, or guaranteed, waterpower in 1853.

James B. Francis, the agent and chief engineer of L&C (see Figure 3), now had to supply 139 11/30 millpowers to 10 textile corporations and a great machine shop. Each of these millpowers represented a specific flow of water in cubic feet per second(cfs) that was expected to produce about 60 net hp on one of the drops in the two-level system (see Figure 4).⁹ Technically, Lowell corporations leased water, not power. Their millpowers should be seen as contracts for the delivery of water with a certain amount of potential energy. The breast wheels or turbines, which converted that energy to mechanical power, did so with varying degrees of efficiency. By installing the latest improvements in turbine technology, mills often increased the net horsepower they generated from each millpower of water that L&C supplied.

Figure 3. James B. Francis, agent and chief engineer of the Proprietors of Locks and Canals. Francis developed methods to deliver, measure, and charge for surplus water in Lowell. He also designed hydraulic turbines and advised textile corporations on the use of water and steam power. Courtesy of the Center for Lowell History, University of Massachusetts-Lowell.

Figure 4. The Lowell Canal System in 1848. This two-level canal system supplied water to 10 textile corporations and a large machine shop. For most of a typical year, the river and the power canals had the capability to provide much more than the minimum flow stipulated in leases. Map drawn by Mark Howland in 1975: "Lowell Canal System, 1848," HAER MA-1, sheet 2 of 2. Courtesy of Historic American Engineering Record.

The millpowers assigned to a corporation were only a guaranteed minimum. Lowell's millpond, which extended for 19 miles above Pawtucket Dam, had to supply 3,595 cfs for 15 hours of every workday to satisfy the leases for 139 11/30 permanent millpowers on two levels. Meeting that requirement was rarely a problem, even in the summer months. The annual flow of the Merrimack fluctuated but was usually close to the modern (1923–1997) average of approximately 7,000 cfs.¹⁰ By storing all the water that came down the river when the mills were closed (nights and Sundays), L&C could theoretically satisfy all its contractual obligations with a river flow of only 1,926 cfs. In all but an extended drought, there was water to spare.

The Proprietors of Locks and Canals took their contractual obligations very seriously. After all, the power company was wholly owned by the manufacturing corporations that leased waterpower. Even when water seemed abundant, there was apparently no serious discussion of increasing the number of permanent millpowers beyond the 139 11/30 set in 1853. The predictions used to set that conservative number were based on careful analysis of meteorological and hydrological data, which turned out to be remarkably accurate. When terrible droughts did occur (as in 1880), Francis had to struggle to supply the minimum flow. He much preferred to offer surplus water most of the year than to lease additional millpowers that he might not be able to deliver at all times.¹¹

According to the indentures (leases) of 1853, manufacturers on the system had the right to use surplus water under terms and conditions that were to be set by L&C. For the next six years, Francis allowed mill superintendents to take additional flow (above the amounts in their leases) without charge, as long as they created no serious problems with the operation of the canal system.¹² The chief engineer, however, had already begun to think of ways for L&C to control the distribution of this surplus and make a profit in the process.

Measuring Flow and Charging for Surplus Water

Francis, who gained international renown with the publication of Lowell Hydraulic Experiments in 1855, had developed sophisticated techniques for measuring the water used by each of the mill complexes on the canal system. By 1856, he perfected a gauging system that relied primarily on hollow metal tubes, weighted with lead to float vertically in the current (see Figure 5). Teams of four to five men timed the passage of these tubes through carefully sized, wood-lined flumes in the canal system and then calculated the flow to particular corporations. In some cases, Francis relied on turbine gate settings to estimate water consumption, but he much preferred the tubes, which gave results that were more accurate. During normal conditions, his busy crews did four measurements per week in most of the canals. Their data helped Francis distribute water fairly and made it possible for L&C to charge manufacturers extra for any surplus they used (see Figure 6)¹³

Figure 5. Meeting of the American Society of Civil Engineers in Lowell in 1878. The engineers are watching a demonstration of James Francis's method for measuring flow through a canal. At the lower left is a man with a long tube that he will float down the Merrimack Canal through the wood-lined section serving as a testing flume. Observers are sitting on the first of two transit beams marked to indicate feet of distance from one end. With frequent flow measurements, Francis could charge corporations for any surplus water that they used. Courtesy of the Lowell Historical Society and the Center for Lowell History, University of Massachusetts-Lowell.

Figure 6. The Eastern Canal at the upper end of the Boott Mills complex. In the canal is a surviving transit beam (next to a pair of beams that once supported a flume measurement bridge). The railway bridge at the right leads to the altered entrance of the Boott Mills coal pocket. In the background is the chimney of the neighboring Massachusetts Mills, which also used a combination of waterpower and steam power. Photo by author.

Confident in his ability to determine how much surplus flow was going to each mill complex, Francis proposed in 1856 that "rents should be paid to this company for such excess." The directors of L&C took no immediate action on his suggestion, but Francis continued to collect data that supported his arguments. Three years later, his investigations showed that "the use of Surplus Power" had been "gradually increasing." He reported, "In 1856, by measurement there were 26.63 [surplus] Mill Powers in use; in 1857, 34.31 Mill Powers, and at present [1859] I think there must be about 40 Mill Powers in use." All of this water was in addition to the permanent leases of 139 11/30 millpowers. The only tool that the chief engineer possessed for regulating surplus at that time was his power to prohibit its use entirely.¹⁴

Francis warned the president and directors of L&C that he needed a better way to restrain the consumption of surplus water. Demand for power was rising, and the capacity of the canals was limited. He wanted to charge for surplus at "a very high rate, as high as it will bear, irrespective of the cost of the permanent powers." After going through the records of L&C, Francis determined that the average cost (considering both the initial purchase price and the annual rental) of a permanent millpower of water in Lowell was about $3.75 per working day. The cost of an equal amount of steam power in the city was much higher by his estimates: $9.71. Francis then asked, "Would half the price of steam power be too much to charge for the surplus water power?" He concluded that $5 per millpower per day was an appropriate fee. The chief engineer also requested authority from the directors to stop "the use of part of the surplus as well as the whole."¹⁵ Francis wanted more flexibility in setting limitations.

In 1859 and 1860, the directors of L&C finally established policies and regulations for renting surplus water. They were careful to retain their "full right of limiting or prohibiting the use of the Surplus Power by any company." The canal company was not conceding any "permanent right or privilege whatever." They wanted to treat the textile corporations fairly and to avoid any sense of favoritism. Any limitation would be the same percentage (25%, 10%, etc.) of each corporation's permanent millpowers. Francis had the authority to change that maximum percentage of surplus water at any time. Whenever there was not enough water for every corporation to have some extra flow, then surplus would be prohibited for all. Manufacturers only had to pay for the additional water they actually used, with billing beginning at $3.50 per millpower per day. To encourage restraint and prevent too much dependence on surplus water, the fee for any excess over 30% of the leased amounts jumped sharply to $7 per millpower per day.¹⁶

The fee structure changed over time. In 1866, the directors doubled the modest fees they had set in 1859. Four years later, they seemed to encourage reasonable use of surplus water by cutting the minimum charge to $5 per millpower per day and the charge at the next level (31%–40% of the leased power) to $10 per millpower per day.¹⁷ They probably believed that falling costs for steam power made such an adjustment necessary. None of the directors wanted to see corporations choosing to run steam engines instead of water turbines. However, restraining water consumption was also on their minds. Their delivery system did not work well when a great deal of surplus water was moving through the canals. They decided to penalize companies taking surplus water equal to more than 40% of their leases. For surplus from 41% to 50%, they set a high fee of $20 per millpower per day. If a corporation exceeded the 50% ceiling, it would have to "pay for the whole amount of surplus power used by it, at the rate of twenty dollars per mill-power per day."¹⁸

By the time of the 1880 census, the $5 fee applied to allowable surplus up to 40% of the leases. It was considered "a contribution toward the annual expenditures of said Proprietors [of Locks and Canals]." The $10 fee was for 41%–50%, and the $20 fee for 51%–60% (or for all of its surplus, if a mill exceeded 60%). There was no mercy for anyone who exceeded a special limit set by the chief engineer when water was scarce: "If during the time when such limitations are in force, any company shall use a quantity of water in excess of that to which it is limited, said company shall pay for the same at the rate of $75 per mill-power per day for the greatest excess used on each day."¹⁹

L&C always provided an important exception in times of natural backwater, when mills along the river had to take more than their leased flow to get the minimum power they needed to operate. Backwater was a condition caused by high water levels in the Merrimack. In even modest freshets, the river flooded wheel pits in the mills that discharged into it. Since this backwater impeded the rotation of breast wheels and reduced the effective head for turbines, mills got less mechanical power (horsepower) from each millpower of water. Only by using more water could they make up the gap. At first, L&C charged nothing for surplus water during backwater conditions (established by a height gauge in the river behind the Merrimack Mills). In 1875, they established a nominal fee of $1 per millpower per day. Although it went up to $2 in 1885, the fee was still far below any of the charges for extra flow when there was no backwater.

Surplus water fees were a sharp departure from the normal practice of purchasing rights to and then paying an annual rent for permanent millpowers. Water had clearly become a commodity when you paid for only what you used on a given day. With permanent millpowers, you paid to take a specified flow of water during every workday.²⁰ If you did not need the contracted water, you would still have to pay the rental charges or risk losing the rights that you had purchased. Surplus water was different: it was an option, available on demand, up to the limits set by L&C. Although L&C, which was wholly owned by the corporations it served, was not in business to make a profit, extra income from surplus water sales was welcome. The annual rent for each of the 139 11/30 permanent millpowers (after paying for the initial leases) was $300 per year. That amounted to a total of $41,810 per year, "about the whole of which is expended in maintaining and improving the water-power."²¹ Surplus fees apparently made possible many of the special projects that enhanced Francis's sterling reputation in the engineering profession. His meticulous work on hydraulic and mechanical experiments continued throughout his career. L&C heavily subsidized the publication of three revised additions of his Lowell Hydraulic Experiments between 1868 and 1883.

Collection of surplus water fees, which had dropped off during the Civil War, picked up with the return to full production in the mid-1860s. Continuing increases in the number of spindles (the principal indicator of manufacturing capacity in the cotton textile industry) created more demand for waterpower in Lowell. Taking regular measurements of flow was an expensive, labor-intensive process, but Francis could justify it by pointing to the money earned by surplus water. A record entry for fiscal year 1881 listed $9,551.70 as the total cost of measuring water. The L&C engineering office subtracted that amount from the $29,152.21 paid by the corporations for extra power. The provision of surplus water netted L&C $19,600.51 during that 12-month period. That was a substantial figure, almost half as much as the income from renting permanent millpowers.²²

There were tradeoffs for this additional money. Taking too much surplus water could create problems for both the canal company and its customers. Heavy demands by just a few mills could upset the dynamic equilibrium in the canal system. All but one complex on the upper level discharged into the lower level. In order to prevent waste, Francis tried to keep the demands on both levels in balance.

Another problem was friction head losses caused by excessive current in the canals. Energy lost to friction was reflected in the diminished head (energy of position). Some of the waterways had worse losses than others did, but all suffered drops in head as water brushed against rough sides and bottoms on its way to the mills. Manufacturers who drew surplus water caused higher currents, and thus more friction, in the canals. Mill superintendents were glad to have the additional flow, but they were dismayed by the diminished head. There were frequent situations in which water arrived 2 to 3 feet lower than the established level.

The higher current also created difficulties for navigation in the canals, which still functioned as transportation corridors as well as conduits for power. Although much less cargo (mostly logs for Lowell's sawmills) moved through the canals than in earlier periods, L&C was still bound by law to maintain the transportation capability. Francis warned in 1858, "the extensive use of surplus power will increase the rapidity of the current to an extent that will expose us to inconvenient interference and may oblige us to limit its use in some localities."²³

Finally, there was the danger of releasing too much water from the New Hampshire reservoirs too soon, leaving an inadequate reserve for a possible drought in the fall. Francis was very careful when he drew down the lakes that fed the Merrimack River. He kept detailed records of lake levels and daily rainfall, but it was even more difficult to predict the weather in New England in the 19th century than it is today. Whenever there was any doubt about maintaining sufficient storage, he put limits on surplus water. The percentage of limitation was the same for each corporation, but it was applied to the number of millpowers in the individual leases, which ranged from 24 2/3 for the Merrimack Manufacturing Company to 3 3/10 for the Lowell Machine Shop (Table 1).²⁴ As mentioned above, financial penalties for exceeding the restrictions were heavy, and L&C could cut off water to any corporation that ignored its orders. Everyone agreed that the highest priority was always assuring enough flow for the permanent leases.

Table 1 A Comparison of Leased and Installed Waterpower in 1883

Mills	Leased Millpowers	Equivalent Leased Horsepower*	Ratio Installed/Leased	Equivalent Installed Horsepower*

Merrimack Hamilton Appleton Lowell Middlesex Suffolk & Tremont† Lawrence Boott Massachusetts Lowell Machine Shop Total	24 2/3 16 8 8/15 8 2/5 5 23/30 13 17 3/10 17 13/15 24 8/15 3 3/10 139 11/30	1,727 1,120 597 588 404 910 1,211 1,251 1,717 231 9,756	2.14 1.76 1.58 0.90 1.34 1.47 2.18 2.04 2.01 2.59	3,695 1,971 944 529 541 1,338 2,640 2,551 3,452 598 18,259
Sources: L & C Director's Records, Dec. 17, 1853, and Appendix, Shedd & Sawyer Reports, 1883, Vo l. A17, Francis File #83, Baker L & C. *For this table, each millpower in 1883 is considered to be equivalent to 70 net horsepower. †The Suffolk and Tremont Mills were merged by this date.

Clemens Herschel, who operated the canal system in Holyoke, Massachusetts, adopted similar rules for providing surplus water. He was a younger associate of Francis and shared his concerns about companies that took too much. At a birthday dinner, a friend reported that Herschel "restricted the use of water at times of low flow and penalized by ten times the regular rate any who exceeded their allotment. An irate customer, after fruitless expostulation, quoted some lines about being 'crowned with a little authority' and making angels weep. Mr. Herschel rejoined that if the angels wept any surplus tears they would have to pay for them."²⁵

Hydraulic Turbines

Great changes in power technology were underway in Lowell when L&C began charging for surplus water in 1859. Most of the corporations had already improved their capability to use waterpower with the purchase of turbines. Some mill superintendents had anticipated the additional power that would result from the canal improvements and the purchase of the New Hampshire lakes. They were also aware of the higher efficiency that turbines offered. When the Civil War began in 1860, only 3 of the 10 textile corporations had any breast wheels left in their mills.²⁶

Companies often installed much more turbine capacity than was necessary to handle the flows guaranteed in their leases with L&C. Turbines could handle a wider range of both head and flow than could breast wheels. The choice of a more powerful turbine might not increase the purchase price a great deal or require a significantly larger wheel pit. Any of the mill superintendents ordering new turbines would seek enough capacity to run all their machinery with a few feet of backwater, and most of them wanted to take advantage of surplus power whenever it was available.

It is difficult to assess the exact power that mills on the system really produced at any point in time, but we do have an engineering report from 1883 that compared the capacity of installed wheels at each corporation with the amounts of water in their permanent leases. Only the Lowell Manufacturing Company, a maker of carpets, had wheels that could not handle the leased flow. Every other manufacturer had excess capacity. The mean ratio of capacity to leased flow was 1.77. Four of the largest corporations (Merrimack, Boott, Lawrence, and Massachusetts) could accept more than twice the contracted amount of water (Table 1).²⁷

The heavy use of surplus water during backwater conditions is apparent in statistics that L&C compiled for the period from 1875 through 1884. Mills suffered from backwater (determined by a gauge behind the Merrimack Mills) an average of 111 out of the 309 working days in every year. The corporations as a group used an average of 60.23 surplus millpowers in backwater, compared with 24.62 surplus millpowers when there was no backwater (198 days).²⁸ The backwater figure of 60.2 millpowers was equal to 43.2% of the total leases. For no backwater, the percentage was 17.7%. The average number of surplus millpowers in use on the canal system, regardless of river conditions, was 37.37 millpowers, or 26.8% of the total leases.

Changes in the river, the canal system, and turbine efficiency had gradually increased the net horsepower equivalent of a millpower from less than 60 to more than 70 by the mid-1870s. Using the latter figure as a conservative estimate of one millpower, we can calculate that from 1875 to 1884, surplus water (when there was no backwater) provided an average of 1,723 hp in addition to the 9,756 hp from the corporations' 139 11/30 permanent millpowers. The total waterpower produced on the system in the 10 years before 1885 averaged almost 11,500 hp in normal conditions. They took much more water in backwater conditions, but it is impossible to tell how much actual power they realized (since backwater could also reduce the net drop, or head, at a mill complex).

The 1880 U.S. Census stated that "during 'backwater' large amounts [of surplus] are used, some companies using up to 75% of what they own [permanent millpowers]."²⁹ Although we know how many millpowers (or how much water) the mills actually used in backwater conditions (from the 1875–1883 data), we do not know what mechanical power they were getting from that water. The level of the backwater varied constantly, and thus the average head (essential for calculating waterpower) is impossible to determine. It is safe to say, however, that backwater was primarily a problem of the lower level mills along the river, and that it seldom halted production after turbines replaced breast wheels. Most mills that used a lot of free or very cheap surplus water during backwater conditions were probably getting more than enough power from it to compensate for any loss of head.

Hybrid (Steam/Water) Power Systems

The program of surplus water sales had many environmental implications. Although a historian is tempted to credit it with reducing the burning of coal in Lowell (by providing waterpower which was a cheaper alternative to steam power), that was not the case. Some corporate decisions to install steam boilers and engines for textile production must have been influenced by the availability of surplus water most of the year. As we have seen above, much of the steam power on the canal system was supplemental, not primary.³⁰ The paradox is that without surplus water, Lowell's textile corporations might not have invested heavily in steam power.

As early as 1839, Patrick Tracy Jackson, one of the founders of Lowell, was suggesting that L&C should provide "sites for mills to be supplied with water when the river is full, say 8–10 months in the year." He never mentioned steam power, but, even in 1839, that would have been an obvious way to run these mills when no surplus water was available. The new mills were not to have any permanent, guaranteed, millpowers of their own. Nothing came of Jackson's idea, but Francis "revived" a similar scheme for a "power mill" in 1854. He was explicit about the mix of power sources in this building, which L&C hoped to lease. The mill was "to be driven by water whenever we have it to spare for that purpose, the remainder of the year say six months to be driven by steam."³¹ Again, the plan was not implemented, probably because the managers of corporations already on the system wanted to reserve surplus water for their own use. They did take advantage of both surplus water and steam power as they expanded their textile production over time.

The purchase of steam engines and the burning of coal (producing air pollution in Lowell and mining wastes in Pennsylvania) increased sharply after the Civil War. Although a number of engines had been in use at the Merrimack Print Works since 1846, the Statistics of Lowell Manufactures provided no data on steam power until January 1866, when it listed a total of 29 engines, with 2,885 hp, on the canal system. By 1871, there were 30 engines with 3,690 hp, and 10 years later, 88 engines with 11,950 hp (Table 2). However, not all of this power was in use on an average workday. Inexpensive surplus water often kept steam engines in reserve.³²

Table 2 Steam Power and Spindles on the Lowell Canal System

Year	Spindles	Steam Engines	Installed Horsepower

1865 1870 1875 1880 1885	437,420 499,806 692,888 703,670 834,100	29 30 67 88 161	2,885 3,690 7,733 11,950 20,600
Source: Statistics of Lowell Manufactures (Lowell, 1866, 1871, 1876, 1881, 1886). Spindles at the textile mills on the system are one measure of productive capacity. These figures do not include the steam engines at the Lowell Machine Shop, a complex which did not make textiles.

Surplus water actually gained in value once corporations could buy efficient steam engines, like those with the improvements made by George Corliss (see Figure 7). That may seem illogical, but not if you understand the demands of cotton textile production. Each of the mill complexes operated with a system of manufacturing that was set up for continuous flow from raw cotton to fabric.³³ The system was most economical when it ran at full capacity. Running only part of the machinery was difficult, since managers could not simply shut down one part of the process for any significant period. Machines at one stage of production fed the next stage, and material stacked up quickly if any part of the system was down. To run at reduced capacity, a mill superintendent would have to keep the system in balance, shutting down some machines in each department and laying off workers throughout the complex. It was hard enough to recruit and keep skilled workers without adding the threat of temporary unemployment. Since a consistent level of surplus waterpower would not be available throughout the year, a new mill that had no supplemental steam power faced high costs for the installation of machinery that would not always be in operation.

Figure 7. A Harris-Corliss steam engine of the type used in Lowell. The William A. Harris Steam Engine Company in Providence, Rhode Island, supplied a number of Lowell corporations with mill engines for hybrid (steam and water) power systems. Courtesy of Slater Mill Historic Site.

In a letter to the president of L&C in 1858, Francis had predicted, "The result of the extensive use of the surplus power at Lowell will be I think, to run it in connection with steam power."³⁴ He was prescient in his recognition that coal-fired steam and surplus water would be mutually supporting. Lowell had a locational disadvantage when it came to steam generation. Coastal cities in Southern New England could get their coal by ship or barge, a less-expensive delivery system than the railroad on which Lowell's manufacturers depended. Costs of rail shipment to interior cities on the Merrimack River were high, even after the completion of the Hoosac Tunnel through the mountains of western Massachusetts in 1875. Railroad bridges did not span the lower Hudson and the mouth of the Connecticut River until the late 1880s.³⁵ It was the availability of inexpensive surplus water that made hybrid (water and steam) power systems in Lowell competitive with 100% steam systems in New Bedford, Fall River, and Providence (see Figure 8).

Figure 8. The Eastern Canal at the Boott Mills in 1910, with chimneys of the Merrimack and the Lawrence mills in the background. The 283-foot smokestack, appropriately named "Jumbo," was the tallest in the United States when added to the Merrimack complex in 1882. Courtesy of the Locks and Canals Collection, Lowell National Historical Park.

It is possible that Lowell's industrial expansion would have stalled without surplus water. No additional permanent millpowers were sold after 1853. Why would you build textile mills in Lowell that required steam power, when you could go to relatively nearby cities where coal was cheaper and other manufacturing costs were similar? If your mill complex could get surplus water at a bargain rate for most of the year, the decision to invest in boilers and steam engines at Lowell made much more sense.

As the chief engineer of the Proprietors of Locks and Canals, Francis advised the textile corporations on both waterpower and steam power, as well as on the interconnected power transmission systems that allowed their combined use. In 1872, John Palfrey, superintendent of the Merrimack Manufacturing Company, was expanding the production capacity in part of his complex, which was the largest on the Lowell canal system. He wrote to Francis: "This increase of machinery will call for another steam engine, and the size & location of this, and the system of main shafting in basements, are matters on which I wish to ask your opinion." Palfrey went on to say, "The power will ordinarily be furnished by one large wheel in #2 Mill, one in #1, & two small wheels in #6, all running with full gates & without regulators, the non-condensing engine running to its full capacity, & the condensing engine furnishing the rest of the power needed for the Yard, & regulating the whole of it." This letter implies that the "wheels" or water turbines were to use as much water as they could take (full gates), while the new regulated (automatically governed) condensing engine would vary its power output (and coal consumption) as necessary to keep mill shafting running at the proper speed for textile production.³⁶

Eventually, some of Lowell's cotton manufacturers installed enough steam capacity to run all their operations without any waterpower at all, but they continued to use their hydraulic turbines as much as possible to reduce coal consumption and thereby minimize their energy costs. When floods created extreme backwater or forced the closing of canals, the interconnected (hybrid) mills did not have to halt much, if any, of their machinery. That security was an added benefit of hybrid systems, which could make use of either steam or water (or both at once).

Valuable Water

Dreams of surplus water probably played a role in the formation of the reservoir system in New Hampshire, and additional flow was a major factor in its continued expansion and maintenance. This radical alteration of both natural and constructed features in the Merrimack River's watershed had widespread cultural and ecological impacts, which this article can only begin to suggest.³⁷ Within minutes of Francis sending a telegraphic message to New Hampshire, a loon floating quietly on Squam Lake (Golden Pond in the popular Hollywood film) would find its elevation changing slowly but steadily as more water was released to feed the mills in Lowell and Lawrence (see Figure 9).

Figure 9. Control structure at the outlet of Squam and Little Squam lakes in New Hampshire. Stone masonry abutments and foundations here date from the Francis era, when northern lakes with more than 100 square miles of surface area were used as reservoirs to store and release water for the mills in Lowell and Lawrence. Photo by author.

The Merrimack River was never dominated, but some of its seasonal flow patterns were altered, reducing the frequency, duration, and seriousness of low water conditions. Surplus water "rents" were a way to profit from natural variations that no engineer could eliminate. They also restrained excessive use of water on a canal system that had built-in limitations.

The men who ran L&C and those who operated its counterpart in Lawrence, the Essex Company, were even more opposed to the "waste" of water over their dams after they began to charge for surplus water. They could not prevent some spillage, particularly in freshets, but any flow that did not go to a wheel pit was water for which they could not bill. This was an important factor in their frequent objections to fishways, which took valuable water from the millponds.³⁸ Their dams hurt not only the upstream migrations of fish like shad, salmon, sturgeon, alewives, and eels but also the downstream movement of both young and adult fish. Shad, which represented a much more important fishery than salmon, and most of which did not die after spawning, had difficulty returning to the ocean.

Shad are easily confused and blocked by obstacles such as dams. Although they might eventually find a spillway or a small breach in the crest of some dams, they are unlikely to leap over dry masonry or flashboards in hopes of finding water downstream.³⁹ In Lowell and Lawrence the goal of the power companies was to contain all the flow of the Merrimack River in their millponds until the power canals could make use of it. During the workday, particularly in the summer and fall, water in the ponds would fall three or more feet, and it might not rise back to the top of the flashboards before the next morning.⁴⁰ The jagged riverbed below the dam in Lowell was often dry for days at a time.

Flashboards increased both the effective height of the dam and the capacity of its millpond. In 1883, L&C added another foot to the two feet of flashboards that had been standard since 1840. Francis said, "The principal effect of the extra flashboards is in keeping up the head in the canals and to store a larger supply for use of Surplus Power."⁴¹ He wanted the pond to fill each night and just begin to spill over the top of the flashboards in the early morning when his men opened the head gates of the canals. To him and his company, any additional spillage was waste (see Figure 10).

Figure 10. Workers replacing wooden flashboards on the masonry dam at Pawtucket Falls in Lowell. Flashboards (supported by iron pins) raised the effective height of the dam, thus ponding more water behind it and increasing its capability to supply surplus water to the canal system. Photo by author.

One of Francis's first assignments after arriving in Lowell was to plug leaks in the Pawtucket Dam, and he continued to worry about water escaping from the pond throughout his career. He was not known as "the chief of police of water" for nothing.⁴² Any waste of water offended his sensibilities. He wanted to provide his corporate customers with as much surplus water as the river and reservoirs could supply and the canal system could handle. He even opposed, but failed to prevent, the development of a municipal water system in Lowell in the 1870s, largely because its pumps would take water from the millpond. Francis was a respected civic leader in Lowell, but he was very protective of his company's water rights. It is clear that the city's water system, which had obvious public health benefits, was a much bigger threat to surplus water sales than it was to the delivery of "permanent," contracted millpowers. The city's one-time payment of $50,000 for water from L&C was not enough to satisfy the chief engineer or his directors, but it helped to offset some of the income they expected to lose from surplus water sales.⁴³

The manufacturing corporations that owned L&C backed Francis when he argued publicly against the city's new water system. The chief engineer and the corporate directors were usually in agreement on political issues that affected industry. There was, however, one issue on which Francis and the textile producers may have disagreed: the 10-hour law of 1874. This law, supported by labor reformers but opposed by most manufacturers in Massachusetts, restricted the length of the workday in factories. A shorter workday would mean less weekly production of fabric, but it would also make it much easier for L&C to meet contractual obligations for water. Although the leases still called for 15 hours of water use, Francis knew that the 10-hour law would reduce the hours that he had to leave his head gates open. He would not only be required to deliver less water in a shorter workday but would also have more time to refill the millpond after the mills closed. His ability to provide surplus water would therefore be much improved. Soon after the law passed, L&C raised the limits from 30% to 40% for surplus water at the minimum charge of $5 per millpower per day. This action encouraged mills to draw extra water when it was plentiful. In 1881, Francis remarked that he had been forced to prohibit all use of surplus water only 40.5 days since 1874. The most extensive restrictions had come in the exceptional drought of 1880.⁴⁴

After Francis retired on 31 December 1884, he continued to consult for L&C and remained interested in surplus water. In 1888, the directors asked him to give his opinion on whether they should sell the northern reservoirs, which legislators and industrial interests in New Hampshire were pressuring them to do. He warned that there would be "a greater irregularity in the supply of water, requiring more frequent changes, than at present, in the limitations to the use of Surplus Power, and in the relative amounts of steam and waterpower used." However, he still favored the sale for a number of reasons. One of them concerned the increasing number of spindles in Lowell every year and the growing importance of steam power in driving that additional machinery:

Water Power cannot be relied upon to operate the increased amount of machinery, except to a comparatively small extent during part of the year. The reliance must be on steam power. To run all the machinery uniformly a plant of steam power must be maintained, sufficient for all emergencies, and also of waterpower, to enable it to be used as far as practicable to reduce the consumption of coal. A large proportion of the waterpowers in New England are already operated on this basis. The result in Lowell will be that irregularities in the supply of water will become of less and less importance, but it will be economy to use it as much as possible.⁴⁵

Conclusion

Surplus water had been an important supplement to the permanent millpowers in Lowell since the 1850s. Historians have overlooked the significance of surplus water, but it explains many of the policies and actions of the Proprietors of Locks and Canals. It provided an important advantage as Lowell struggled to compete with growing textile centers in southeastern New England, and it encouraged the installation of steam engines and hybrid power systems.

Generally, water shortages are considered to have provided the greatest incentives for the introduction of steam power at waterpower sites. The Lowell case suggests that scholars should not accept this characterization uncritically, for at Lowell it was not water shortages that played the greatest role in the adoption of steam power but water surpluses. While the present study has focused on Lowell, there is evidence that this may also have been the case at other great American waterpower centers like Holyoke and Lawrence.⁴⁶

Material evidence from the wheel room of the Suffolk Mill prompted the archival investigations that demonstrated the importance of surplus water and its links to steam power. Further examination of the industrial landscape of Lowell has confirmed that Lowell mills used more waterpower than previously assumed, that precise measurement of flow was a regular part of system operations, and that hybrid (steam/water) power systems lasted well into the 20th century. The need for surplus water may have declined by the 1890s, but it remained an important asset that could cut coal consumption. Power from surplus water in Lowell was always cheaper than steam power. In the textile industry, every cost savings counted (see Figure 11).

Figure 11. Chimneys and coal smoke above textile mills on the Lowell Canal System. This photograph, taken in 1910 from the top of the new smokestack of the Massachusetts mills, looks upstream with the Merrimack River on the right. Chimneys at the Massachusetts, Boott, Merrimack, Suffolk, Tremont, and Lawrence mills are visible, and most are in use. Courtesy of the Locks and Canals Collection, Lowell National Historical Park.

Acknowledgements

I would like to thank the many people who have assisted me in my research on Lowell over the years. Charles Parrott not only shared ideas and information but also commented on several drafts of this article. Particularly helpful on this project were Ann Booth, Bob Brugger, Gray Fitzsimons, Mike Folsom, Marti Frank, Greg Galer, Bob Gordon, Carolyn Goldstein, Rick Greenwood, Larry Gross, Duncan Hay, Mark Herlihy, Charles Hyde, Larry Lankton, Ed Layton, Steve Lubar, Lyn Malone, Pat Martin, Martha Mayo, Lance Metz, Mike Raber, Terry Reynolds, Matt Roth, Giovi Roz, Clare Sheridan, George Smith, Joel Tarr, Dan Walsh, Claiborne Walthall, the American Museum of Textile History, Baker Library of Harvard University, Brown University (including Libraries and Instructional Technology Group), Burndy Library, the Center for Lowell History of the University of Massachusetts at Lowell, Dibner Institute, Historic American Engineering Record, Johns Hopkins University Press, Lowell Historical Society, Lowell National Historical Park, the Society for Industrial Archeology, and the Tsongas Industrial History Center. This article draws heavily on a paper that I gave at an international conference on rivers, organized by the German Historical Institute in 2003. In my recent research on industrial steam power, I have worked closely with Marti Frank, who is writing a dissertation dealing with that topic at Harvard University. I am also grateful to the anonymous reviewers for IA, who made very helpful suggestions.

Notes

^1. The historians on that team were the author, Charles Parrott, and Charles Hyde.

^2. We later found an insurance drawing in the counting house that showed the seven Victor turbines (installed in 1895) and gave their total capacity as 1,683 hp. This was far more than the 455 hp one could expect to get from the Suffolk Mills' 6.5 millpowers of leased waterpower (at 70 hp per millpower). Although some of these wheels may have been intended as backups, their capacity was far greater than we expected. The Wannalancit Company that occupied the site in the 1970s had insurance surveys from 1910, 1913, and 1915.

^3. The teachers were part of an institute, or workshop, run by the Tsongas Industrial History Center, which asks the author to give lectures and lead tours one or two times per year.

^4. W. Kenneth Hamblin, The Earth's Dynamic Systems (Minneapolis, Minn.: Burgess, 1985), 20–22, 157–60; Robert B. Gordon, "Hydrological Science and the Development of Water Power for Manufacturing," Technology and Culture 26 (1985): 204–35.

^5. See J. W. Meader, The Merrimack River: Its Source and Its Tributaries (Boston: B. B. Russell, 1869), 40, 46–47; George Swain, "Water-Power of Eastern New England," Tenth Census of the United States, vol. 16 (Washington, DC: GPO, 1885), 71–72, 98–104.

^6. Henry A. Miles, Lowell As It Was, and As It Is (Lowell, Mass.: Nathaniel Dayton, 1845), 14.

^7. The Proprietors of Locks and Canals had opened the Pawtucket Canal around the falls in 1797, but it was never a financial success. See Patrick Malone, Canals and Industry: Engineering in Lowell, 1821–1880 (Lowell, Mass.: Lowell Museum, 1983). This article is reprinted with other contributions in Robert Weible, ed., The Continuing Revolution: A History of Lowell, Massachusetts (Lowell, Mass.: The Lowell Historical Society, 1991), 137–55. See also Louis Hunter, Waterpower, vol. 1 of A History of Industrial Power in the United States (Charlottesville: Univ. Press of Virginia, 1979); Theodore Steinberg, Nature Incorporated: Industrialization and the Waters of New England (New York: Cambridge Univ. Press, 1991); Edwin Layton, From Rule of Thumb to Scientific Engineering: James B. Francis and the Invention of the Hydraulic Turbine (Cambridge, Mass.: MIT Press and McGraw Hill, 1992); and Patrick Malone and Charles Parrott, "Greenways in the Industrial City: Parks and Promenades along the Lowell Canals," IA: The Journal of the Society for Industrial Archeology 24, no. 1 (1998): 19–40.

^8. Steinberg, Nature Incorporated, 102–13 (see n. 7).

^9. Patrick Malone, "James B. Francis and the Northern Canal," in Boston's Water Resource Development: Past, Present, and Future, ed. Jonathan French (New York: American Society of Civil Engineers, 1986), 10–18; James B. Francis, "The Standard Mill Power at Lowell," memo of 7 Aug. 1858, vol. A18, file no. 90, Locks & Canals Papers, Baker Library at Harvard University (henceforth abbreviated as Baker L&C); Locks & Canals (henceforth abbreviated as L&C) Directors' Records, 17 Dec. 1853. A typescript of the Directors' Records is in L&C Collections, Lowell National Historical Park (henceforth abbreviated as LNHP), with microfilm copies at the Baker Library and at the Center for Lowell History at the University of Massachusetts at Lowell (henceforth abbreviated as CLH).

^10. L&C Directors' Records, 17 Dec. 1853 (see n. 9); Water Resources Data Massachusetts and Rhode Island Water Year 1997 (Washington, DC: U.S. Geological Survey, 1998), 54–57. For the flow rate at Lowell, you have to adjust the modern USGS data for the Merrimack River to allow for the location of gauges. You must subtract the Concord River flow that enters the Merrimack below Lowell's Pawtucket Dam but just above the Merrimack River gauging station.

^11. An interesting question raised by one referee is whether financial considerations played a role in the retention of the limit on permanent millpowers, despite experience showing that extra water was almost always available. There might have been greater profit potential for L&C in the rental of surplus water than in the lease of additional permanent millpowers. However, payments for permanent leases were assured (if the minimum flow was provided), while surplus profits depended on voluntary rentals. In an economic depression, for instance, mills were under no obligation to use (and pay for) any surplus. The manufacturing corporations also owned L&C and ran it largely for their own benefit. Excess profit at L&C would come from payments by the corporations. James B. Francis might get to spend some of the profit for his own pet projects (experiments, landscaping, etc.), but he also had to keep his employers happy. The author believes that Francis's greatest concern was meeting contractual obligations. He may have been overly conservative, but droughts did occur, and after the city started taking river water for its urban needs in 1876, he had slightly less flow with which to provide permanent millpowers.

^12. L&C Directors' Records, 12 April 1859 (see n. 9); Form of Lease, 1853, Sundry Papers volume, L&C Collections, LNHP (see n. 9).

^13. Patrick Malone, "'A Good Guide for the Engineer': James B. Francis and Flow Measurement in Lowell," News from the Burndy Library 4, no. 1 (Winter 1997): 16–20; Swain, "Water-Power," 82–83 (see n. 5). Charles Parrott gave a detailed paper, entitled "'The Chief of Police of Water': James B. Francis and the Oversight of Waterpower Distribution in Lowell," on the tube method of flow measurement, presented at the SIA's May 1997 conference in Houghton, Michigan [copy in author's possession]. Parrott and the author are writing an article together on this topic. Parrott is also working on the topic of tall chimneys, which relate to steam power development.

^14. L&C Directors' Records, 5 Nov. 1856, 21 March 1859, 12 April 1859 (see n. 9). A depression in 1858 reduced production so much that James Francis did no measurements that year.

^15. JBF to Thomas Carey, 6 Dec. 1858, vol. DA-5, Baker L&C (see n. 9).

^16. L&C Directors' Records, 12 April 1859, 11 June 1860 (see n. 9).

^17. JBF to Thomas Carey, 6 Dec. 1858, vol. DA-5, Baker L&C (see n. 9); L&C Directors' Records, 12 April 1859 (see n. 9).

^18. L&C Directors' Records, 17 Feb. 1866, 5 Nov. 1868, 1 Dec. 1870 (see n. 9).

^19. Swain, "Water-Power," 81 (see n. 5).

^20. Steinberg, Nature Incorporated, 85–88 (see n. 7), sees the selling of permanent millpowers (often severed from land sales by the 1830s), combined with the payment of an annual rental fee for the right to use those millpowers, as an important example of the "commodification of water." Payments for the surplus water actually taken (per day) are an even more dramatic illustration of this commodity concept.

^21. JBF to Abbott Lawrence, 2 July 1859, JBF file no. 110, vol. A21, Baker L&C (see n. 9).

^22. Records, 1876–1882, 1 Aug. 1881, 187, L&C Collections, LNHP (see n. 9).

^23. JBF to Thomas Carey, 6 Dec. 1858, vol. DA-5, Baker L&C (see n. 9).

^24. L&C Directors' Records, 17 Dec. 1853, JBF statement of millpowers (see n. 9).

^25.Engineering News-Record (29 Mar. 1928). Herschel was then 86.

^26.Statistics of Lowell Manufactures (annual broadside, Lowell: 1859–1861) [copy in author's possession].

^27. Appendix, Shedd & Sawyer Reports, 1883, vol. A17, file no. 83, Baker L&C (see n. 9). The heading of this table is incorrect, but one can see from the cfs data that this is the ratio of capacity to leased amounts.

^28. Records, vol. 10a (1884–1891), 18, L&C Collections, LNHP (see n. 9).

^29. Swain, "Water-Power," 81 (see n. 5).

^30. Supplemental or auxiliary steam power had long been used in combination with waterpower for manufacturing. In Britain, there are examples in the 1740s and 1750s of Newcomen steam engines pumping water back up to vertical waterwheels that produced rotary power. This practice could artificially augment or replace stream flow in droughts, provide a continuous source of additional power for expanded production, or function in a closed system not connected to a stream. Later in the century, engines by James Watt, John Smeaton, and others (including Joshua Wrigley's modifications of the1698 Savery engine) joined Newcomen engines in pumping for waterwheels. However, this type of steam/water combination became obsolete after Watt's improvements in the rotative engine between 1780 and 1800. In mills of the 19th century, steam engines could assist, fill in for, or completely replace waterwheels and turbines. The use of auxiliary steam engines became a common practice in both Britain, where available waterpower was limited, and in the United States, which had numerous falls with relatively steady and abundant flow. Such hybrid systems were also used in France but to a much lesser degree. The Wilkinson Mill (a textile mill with a machine shop) in Pawtucket, Rhode Island, had both a steam engine and a breast wheel when it opened in 1811. In the American textile industry there were many water mills, steam mills, and combined steam/water mills. The combined or hybrid power systems deserve much more study than they have received. Although many of these systems were designed to handle occasional shortages of waterpower (during droughts, ice conditions, or floods), they also provided exhaust steam for heat and processing, allowed manufacturers to enlarge mill complexes, and worked particularly well at sites that had surplus waterpower for most but not all of the year. One of the problems with the available data on hybrid power systems is that we may know the horsepower of installed steam engines but not know how often or how hard they were run. That makes it difficult to compare the use of waterpower and steam power at many sites. See Terry Reynolds, Stronger Than a Hundred Men: A History of the Vertical Waterwheel (Baltimore, Md.: Johns Hopkins Univ. Press, 1983), 321–31; Richard Hills, Power in the Industrial Revolution (Manchester, England: Manchester Univ. Press, 1970), 93–94, 102, 108, 134–40; 161–76; Neil Cossons, The BP Book of Industrial Archaeology (Newton Abbot, England: David & Charles, 1973), 70, 91–92; Hunter, Waterpower, 286–87, 498, 514–29 (see n. 7); Louis Hunter, Steampower, vol. 2 of A History of Industrial Power in the United States (Charlottesville, Va.: Univ. Press of Virginia, 1985), 110; Carroll Pursell, Early Stationary Steam Engines in America (Washington, DC: Smithsonian Institution Press, 1969), 84–89, 134; Gary Kulik and Patrick Malone, The Wilkinson Mill: A National Historic Mechanical Engineering Landmark (Pawtucket, R.I.: ASME and Slater Mill Historic Site, 1977), 2–4.

^31. P. T. Jackson to Directors of L&C, 13 Sept. 1839, vol. A1, Baker L&C (see n. 9); JBF to Corliss & Nightingale [steam engine builders in Providence, R.I.], 1 Jan. 1854, file no. 14, vol. A2, Baker L&C (see n. 9).

^32.Handbook for the Visitor to Lowell (Lowell, Mass.: D. Bixby, 1848), 11; Statistics of Lowell Manufactures (Lowell, Mass.: 1866, 1871, and 1881). The Merrimack Print Works could use exhaust steam for heat in its finishing processes, one reason for the early adoption of steam power in that part of the vast Merrimack Mills complex. Steam, whether as engine exhaust or produced simply for heat, was also considered the best way to warm working spaces in the city's mills by 1850. There was an increase in steam power installation during the Civil War years, when most of the textile corporations in Lowell sharply limited or ceased manufacturing. Many of them focused instead on improvement or expansion of their mill complexes. Installing engines or planning for future use of steam power was part of some wartime projects. Because of the production cutbacks, Lowell published no annual statistics between January 1861 and January 1866. The January 1866 Statistics of Lowell Manufactures was the first to include tabular data on steam engines and installed steam horsepower. During the war, the Suffolk Mfg. Company installed an engine, and the neighboring Tremont Mfg. Co. built a boiler house and smokestack with sufficient capacity to handle the needs of an engine when the company finally decided to add one. See Ann Booth's report, "Tremont Yard," for the Lowell National Historical Park, Lowell, Mich., n.d., pp. 14, 27. For discussion of another mill complex that used a hybrid power system, see Laurance Gross, The Course of Industrial Decline: The Boott Cotton Mills of Lowell, Massachusetts, 1835–1955 (Baltimore, Md.: Johns Hopkins Univ. Press, 1993), 42–43, 54–56.

^33. Richard Greenwood and Patrick Malone, "The Mill As a System," report from The Center for History Now to Lowell National Historical Park, 28 Oct. 1983, Lowell, Mass.; Robert Gordon and Patrick Malone, The Texture of Industry: An Archaeological View of the Industrialization of North America (New York: Oxford Univ. Press, 1994), 348, 368–69.

^34. JBF to Thomas Carey, 6 Dec. 1858, vol. DA-5, Baker L&C (see n. 9).

^35. Larry Lowenthal, From the Coalfields to the Hudson (Fleischmanns, N.Y.: Purple Mountain Press, 1997), 111–13, 131; Ronald Karr, The Rail Lines of Southern New England (Pepperell, Mass.: Branch Line Press, 1994), 158. The Hoosac Tunnel opened a relatively direct route from Boston to the upper Hudson in 1875; the Poughkeepsie Railroad Bridge spanned the lower Hudson in 1888; and the shoreline route on Long Island Sound was completed over the Connecticut River in 1889.

^36. John Palfrey to JBF, 25 June 1872,"Estimate of power shafting, etc. for Merrimack Mfg. Co. July 1872," file no. 54, vol. A13, Baker L&C (see n. 9). I am grateful to Marti Frank of Harvard University who found this letter during her dissertation research and shared it with me. Her scholarly contributions have enhanced this article.

^37. The best coverage of the northern reservoirs and their impact is in Steinberg, Nature Incorporated (see n. 7). The author of this article is now completing a book on waterpower in Lowell for Johns Hopkins Univ. Press. That manuscript has extensive coverage of environmental issues in Lowell and in the watershed of the Merrimack River.

^38. Steinberg, Nature Incorporated, 189–203 (see n. 7).

^39. John McPhee, The Founding Fish (New York: Farrar, Strauss, and Giroux, 2003), 29–32, 95–98, 249–58; Records H, 112, L&C Collections, LNHP (see n. 9).

^40. See for example, Day Book no. 15, 16 Sept. 1884, 208, L&C Collections, LNHP (n. 9).

^41. JBF to Directors of L&C, 17 Aug. 1888, Legal Opinions volume, L&C Collections, CLH (see n. 9).

^42. Records H, 1866–1879, 120, L&C Collections, LNHP (see n. 9); William Worthen, "Life and Works of James B. Francis, Contributions to the Old Residents Historical Association, vol. 5 (Lowell, Mass.: Old Residents Historical Assoc., 1894): 234.

^43. JBF to Directors of L&C, 6 Jan. 1876 and JBF to John Morse, 6 Jan. 1876, vol. DB-8, 366–67, Baker L&C (see n. 9). See also Jim Beauchesne, The Water Question, or, Whose River Is This? memo prepared for Michael Wurm, 16 Sept. 1994, LNHP (n. 9); and Claiborne Walthall, "Power vs. People: City and Corporation Relations in Nineteenth Century Lowell, Massachusetts," (honors thesis, Urban Studies Program, Brown University, Providence, R.I., 2002). By giving in to the city's demand for river water, L&C avoided a serious political confrontation on both the local and state level.

^44. JBF to Hiram Mills, 24 Sept. 1872, file no. 110, vol. A21, and statement on 1874 amendment to the indentures, file no. 467, vol. 85, Baker L&C (see n. 9); JBF, Report on Improvements, 24 Feb. 1881, Sundry Papers volume, L&C Collections, LNHP (see n. 9).

^45. JBF to Directors of L&C, 17 Aug. 1888, Legal Papers volume, L&C Collections, CLH (see n. 9). The Lake Company, which had controlled the New Hampshire lakes for L&C and the Essex Company, was dismantled in 1889. Steinberg, Nature Incorporated, 269 (see n. 7).

^46. One of the referees suggested most of this paragraph. In 1986, Duncan Hay argued, "The uncertainty of surplus power, even though it was usually limited for only a few days each year, accelerated introduction of auxiliary steam power at Lowell and Lawrence during the latter half of the century." Lawrence had even more surplus power than Lowell. Duncan Hay, "Building 'The New City on the Merrimack,' The Essex Company and Its Role in the Creation of Lawrence, Massachusetts," (PhD dissertation, History Department, Univ. of Delaware, Newark, Del., 1986). See also Swain, "Water-Power," 71–75, 81–82 (n. 5). It is unlikely that any other waterpower company conducted the frequent and precise measurements of water usage that engineers in Lowell and Lawrence used to charge for surplus. However, considerable variation in flow over the course of a year was expected at most sites. Many mills took advantage of surplus water (above low season levels) whenever it was available and used steam engines to supplement that unreliable power source. Even when a waterpower company was in charge at a site, it might allow reasonable use of surplus water without any additional fees.

Content in the History Cooperative database is intended for personal, noncommercial use only. You may not reproduce, publish, distribute, transmit, participate in the transfer or sale of, modify, create derivative works from, display, or in any way exploit the History Cooperative database in whole or in part without the written permission of the copyright holder.