The Fiske Center Blog

Weblog for the Fiske Center for Archaeological Research at the University of Massachusetts Boston.

January 14, 2019
by John Steinberg

Results of Geophysical survey at Langone Park: 100 Years since the Great Molasses Flood

By John M. Steinberg & Grace E. Bello

View of softball diamond at Langone Park.  The infield  and pitcher’s mound is outlined in brown, which helps to orient the following geophysical images.

As we said in our previous post,  the City Archaeologist, Joe Bagley asked us at the Fiske Center if we could conduct a geophysical survey over the area of Langone Park that, 100 years ago, had a tank which ruptured and caused The Great Molasses Flood of 1919.  This is in preparation for Tuesday January 15th, 2019—the 100th anniversary of the disaster.

Grace Bello beginning to set up the geophysical grid by placing PVC flags along the first base line.

For our archaeological geophysical survey, we used two common techniques: ground penetrating radar (GPR) and electromagnetic (EM) conductivity.  The area was first prepared for the survey by placing out a grid of PVC flags.  Because the grid was oriented to the softball diamond, the locations of each of the corner flags (and many of the intermediate flags) were recorded with a survey grade GPS.

John Steinberg walking along the third base line with the CMD-Mini.

First, John used the CMD-MiniExplorer conductivity meter, which requires the operator to walk across the target area holding the unit just above the ground.  These transects are then combined to create a  conductivity map of the subsurface. The CMD-Mini creates a data set with two components at three different “depths.”   The different depths are from the three different receivers in the orange tube at various distances from the single transmitter.  The farther apart the receiver from the transmitter, the “deeper” the reading (1 is the closest, 3 is the farthest and thus deepest).   The two components are complex.  The Quadrature component, usually called bulk conductivity (Con) represents the apparent conductivity of the volume of earth under the unit and is measured in milli-Siemens per meter (mS/m).   Good conductors (e.g., salty wet earth) have high conductivity, while poor conductors (e.g., rocks) have low conductivity.  The in-phase component (IP) is usually expressed in parts per thousand (ppt) and is very sensitive to buried metal.  Thus, we have a total of  six different maps from the CMD-Mini: Con1 & IP1, Con2 & IP2, and Con3 & IP3.

Grace Bello with the Malå GPR unit.

Second,  Grace and John walked back and forth dragging the Malå GPR unit with a 500 MHz antenna.  The GPR unit sends out microwaves and if there is a change in soil moisture (or some other similar property) some of the microwave energy will be reflected back up to the GPR unit, which also has a receiver.  The GPR unit then, like the CMD, collects data along transects which produce a data stream called a radargram. Multiple transects are combined and then sliced at different depths, which allows us to create a series of maps that depict some of the aspects of the changes in the subsurface at different depths.   We created 25 different slices, but only present two below.

Example radargram from the transect 5 m (16.5 ft) north of the third base line.

Outlines of  all the structures from the maps described in our previous blog post

In our recent blog post, we describe the georeferencing of various historic maps of Langone park.  When all of the various structures depicted on these maps are combined, you can get a sense of just how complex this lot is.  Many of the structures may be the same structure, but  with slightly different locations provided by different maps, and we do not know how accurate any of them are.  In this case, we have been asked to identify one of the last structures on the lot.  Generally, the construction of  later structures compromises or destroys the earlier structures.  Thus, our most likely potential target is an area where there is a broad, consistent absence of distinct structures. Furthermore, given the hasty construction of the tank, any remements are probably shallow.  This approach is in stark contrast to our usual method—where we are trying to identify remnants of the earliest structures.

GPR slice 50 cm (about 20 in) below the ground surface.

GPR slice 150 cm  (about 60 in) below the ground surface.

Starting with the GPR results, there is a distinct hard reflector 50 cm below the ground surface (bgs). This hard reflector is circled in pink.  This infield hard reflector is distinct from the outer edge of the infield (marked in brown).  This hard reflector is potentially caused by the remnants of the tank in question.  That being said, we want to be a little careful, because this hard reflector is almost the same as the grass infield area from when this was a little league diamond.  However, this slice is a little too deep to show that contrast.  Furthermore, the wide dirt path from the mound to home plate is not visible in this slice.  Both of these lines of evidence suggest that this hard reflector is a result of current or recent landscaping. The deepest GPR slices do not seem to show remnants the tank but instead show some of the potential dock and landfill boundaries, just to the north of first base.  Interestingly, this dominant southeast angle does not reflect any of the structures or orientations seen in our georeferenced maps.

In-phase for the most shallow CMD-Mini readings (IP1). The potential tank location is in pink.

In-phase for the middle  CMD-Mini readings (IP2).

The CMD-Mini yields much more complex results that correspond to many of the structures outlined in the georeferenced maps.  Starting with the IP components, IP1 shows a blue (high IP) area in the infield that corresponds to the area identified in the GPR (again circled in pink).  Just north of the first base line, in right field, is a rectangular blue area that has the same general dimensions and orientation as the structure seen in the 1917 map, just north of the tank in question.  That structure has an add on (in brown) that curves along the curve of the tank that touches 1st base. The potential tank area is more distinct in the deeper components (IP2 & IP3), while the building in right field is less distinct.

Conductivity for the most shallow CMD-Mini Readings (Con1).

The bulk conductivity component of the data from the CMD-Mini is much more complex, but all three sensor-transmitter distances show the same basic map.

In-phase for the deepest CMD-Mini readings (IP3). The potential tank location is in pink

Conductivity for the middle CMD-Mini Readings (Con2).

The Slatter 1852 Map with Con3 superimposed.

The Bromley 1917 map with Con3 superimposed.

Unlike the IP, the Con does not hint at the tank location, there are three high (blue) conductivity areas that seem to correspond to the distribution and orientation of structures in some of  the georeferenced maps.  The blue area in left center field matches quite closely with the angled structure drawn in the 1852 Slatter map.  Some of the low conductivity lines (which could be lines of rocks) correspond to lines drawn in that 1852 map.  Once a property orientation is established, it tends to persist.  Thus, it is not surprising that the geophysics can correspond to more than one map.  Specifically, the three blue areas in the outfield roughly correspond to the three drawn structures abutting the tank in question depicted in the 1917 Bromley map.

Proposed location of tank superimposed on air photo.

While in both the 1852 and 1917 maps the correspondences with geophysical readings and drawn structures  are not exact, they are well within the range of accuracy that we usually see with these kinds of maps.

When all of our data is combined (the georeferenced maps, the GPR and the Electromagnetic Conductivity) and tried to make fit, our best guess as to the specific location of the remnants of the tank that caused the Great Molasses Flood, is 3 meters northwest of the location drawn in the 1917 Bromley map—at least by our georeferencing of that map.  Obviously, we would need to excavate this dynamic and interesting area to begin to refine the location further, but the geophysical results suggest that the 1917 map is generally accurate.  There is no evidence of a consistent bias in the locations of structures as compared to the geophysics, so as georeferenced, the 1917 map is accurate to better than 5 m (16 ft).  As always, more research is necessary.

The 1917 Broomly map with the proposed actual location of the tank in pink.



Langone Park and Great Molasses Flood of 1919

January 13, 2019 by gracebello001 | 0 comments

By Grace E. Bello & John M. Steinberg

Areal image of what the waterfront looks like today.

The City Archaeologist, Joe Bagley, asked us at the Fiske Center if we could conduct a geophysical survey over the area of Langone Park that, 100 years ago, had a tank which ruptured and caused The Great Molasses Flood of 1919.  This is in preparation for Tuesday January 15th, the 100th anniversary of the disaster.

For our archaeological geophysical survey we used two common techniques: ground penetrating radar (GPR) and electromagnetic (EM) conductivity.   The results will be presented in the next post.

Before we interpret the results of a geophysical survey, we try and georeference every map we can of the area in question.  The georeferenced maps allows us to better understand the geophysical results.  Thus, below is a sample of some of the many maps that we looked at, to understand the complex history of this park.  Today, the park has a little league / softball diamond, and the outline of that feature is shown in brown on each of the georeferenced maps (which are mostly from the Norman B. Leventhal Map & Education Center at the Boston Public Library).  In  great article in the Boston Globe Magazine, you can see a georeferenced 1917 map showing that the location of the tank is in the general area of the diamond.   However, the Wikipedia map shows a smaller tank more north and a building between the tank and Commercial street.  Joe Bagley wanted us to see if we could refine the location of the tank that ruptured in 1919. Over time, maps generally become more accurate, but just how accurate, and how well out team has georeferenced them is and issue that we study intently.  Geophysical results can sometimes help us better georeference the maps and understand what parts are accurate, and what parts might require a little change.   This is akin to the work we have been doing in Plymouth, combining geophysics and georeferenced maps,  getting ready for the 400th anniversary of that colony.

1775 Map depicting the shipyard that occupied the area where the tank collapse

The landscape of Boston has been altered heavily over the past four hundred years.  The evolution of Boston’s landscape is evident in historic maps that depict the city’s waterfront property.  Maps, such as the ones shown in this sequence, are extraordinarily important sources of data about the past. These maps detail the dynamic history of the North End in Boston.

The Great Molasses Flood of 1919, took place in a relatively small portion of the North End’s extensive history and has hardly left a mark on the landscape. The flood occurred on January 15th, 1919, when a 50 ft. tall and 90 ft. in diameter molasses storage tank, owned by the Purity Distilling Company, collapsed.  The collapsed tank then tipped over creating a wave of molasses close to 25 ft. tall which killed 21 people and injured up to 150. Today this is what the area looks like.

1814 Map.  surveyed by J.G. Hales ; J.R. & Penniman.

Starting with the 1722 Captain John Bonner map of the, then, town of Boston, which shows the then active docks that occupied much of the Boston shoreline.  This map was a monumental beginning to modern cartographic detailing that documented Boston’s shoreline. The docks, that are clear in the 1722 map, suggesting  dry and wet docks in the 18th century.  In a redrawn 1775 map, the area where the tank was located in the 20th century, was occupied by Hunts and Whites Ship yard.  In this georeference, the park diamond is centered on the “ar” of Ship Yard.

In an 1814 map, Commercial Street was labeled Lynn Street and the softball diamond seems to be on the mostly dry land with a building to the east.  According to an

1852 Map surveyed and drawn by J. Slatter & B. Callan.

1852 map the adjacent street is now Commercial Street and a substantial amount of land has been added in the form of docks.   It is not clear if the building along the first base line is the same one depicted in the 1814 map, but certainly the building curving around 3rd base seems to be new. The two border lines

1861 map made by Boston City Council, city engineer James Slade

coming together at the pitching mound may be visible in some of the geophysics, and will be seen in many of the later maps.

The same basic configuration is also visible in the 1861 map, made by the City.

1867 Sanborn Insurance map of Boston : volume 1 : plate 1
Cartographer: Daniel Alfred Sanborn

An 1867 map is the first real detailed map of the area. It is also one of the early famous Sanborn fire insurance maps.  It shows several coal sheds and hints that the building along the first base line, first seen in the 1814 map, might still be there 50 years later.

George Washington Bromley 1890 map

The  1890 Bromley map suggests that the “V” of two docks,  meeting at the pitchers mound, first seen in the 1852 map, is still present but that the building  that used to be along the first base line, is now substantially larger.  The docks are also very different.  A image of these kinds of docks can be seen in the birds eye view of downtown Boston dated 1877.  These docks were an active hot spot for landscape change during

Bachman 1877 birds eye view

the 18th and 19th centuries.  Further construction of larger docks and land masses were very common during these centuries by building cribbing or sinking ships to artificially fill in the harbor

By the 1908 map, the lot will take on the

1908 George Washington Bromley map

basic shape, the distilling company is labeled as the owner, and a small tank is indicated in the middle of the lot.

1912 George Washington Bromley map

This may be where the Wikipedia map gets its layout.  In this initial 1908 map one large building can be seen in pink that is replaced by multiple buildings in the following 1912 map.  The 1912 map shows a similar small tank but smaller building on Commercial St, and tracks going into the adjacent lot. To the right of the new park area is the soon to be location of the Purity Distilling Company which seems to have built larger molasses tanks as the business grew. This map shows the first distillery’s tank with two new buildings along side a new above ground railway.

1917 Bromley Map probably showing the tank before destruction.

A 1917 map shows the distillery buildings and the large molasses tank before the accident.   The 1917 map is what we will base much of our interpretation on.  It almost surely shows the large molasses tank that burst during the 1919 disaster as well as two new buildings.  This is the final map that the distillery appears on.  A 1922 map shows the absence of a distillery and tank, replaced with reused buildings by railway operations. Our final map, from 1938, again shows all of the buildings in the 1917 map, with the addition of a small shed where the pitcher’s mound is today.  In 1973 Langone park  was created.

1922 Bromley map

1938 Bromley map

In our next installment we will present the geophysical results from Langone Park.


November 27, 2018
by gracebello001

Updating Cape Cod 3-D Model

Our work with the Cape Cod National Sea Shore  continues as we monitor erosion trajectories.  This 3-D model was built to aid in quantifying beach erosion over time. While the data is collected for purely scientific reasons, I made a fly through movie using photos and GPS control points taken by John Schoenfelder, John Steinberg, Melissa Ritchey, and Jocelyn Lee.

June 15, 2018
by elizabethquinlan002

Chasing Color Changes at Hassanamesit Woods

A view of the summer sky above Hassanamesit Woods

The first few weeks out in Hassanamesit Woods have been marked by (mostly) great weather and even better field experiences. Despite a rain day spent in the lab cleaning recovered artifacts on Monday the 4th, the second week of work gave the students a look at how changing stratigraphy within a unit can both puzzle and inform an excavator. Graduate students Melissa and Liz, joined by graduate student Ivana, began seeing some interesting soil changes as they brought their unit down to roughly 35-45cm below datum. These stratigraphic changes continued as they  followed the strata down to a final depth of about 75cm below datum. As this unit is located right up against the Augustus Salisbury foundation, it was hoped that these soil changes might indicate a builder’s trench in the unit.

View of the Northwest corner of Unit E448 N274

The northwest profile pictured above shows the bands of color that indicate stratigraphic changes. The C horizon is characterized by the greenish-grey sandy layer in the middle.

During construction of a building with a stone foundation it was often the case that builders would dig down into the sterile subsoil (known in this case as the ‘C horizon’ or ‘C strata’) in order to lay the foundation well below the contemporary ground surface. The soil displaced from this digging would then be loosely filled back in, along with building debris and other trash from the time period, so it could be dug out again later if repairs to the foundation were needed. The soil is often put back “out of order”, and areas of clear disturbance in the natural stratigraphy can clue in archaeologists to construction activities at a site. These trenches, and the artifacts recovered from them, can also help date the completion of a foundation.

By the beginning of the 3rd week it became clear that the stratigraphic changes observed in Melissa, Ivana and Liz’s unit were not being found in Rick and Alex’s adjacent unit, meaning the changes must be associated with the foundation rather than the wider yard space. Over in Tyler and Andrew’s unit, however, they began spotting some unusual stone placements, which also continued into Rick and Alex’s unit. At first it was thought they must have been placed deliberately by people in the area, as they were almost all propped up in an ‘upright’ manner. However, upon discussion with environmental archaeologist Dr. Trigg, Dr. Mrozowski, and the discovery of a large amount of loose frost fractured stones, it was decided that they most likely were the result of New England’s at-times violent freeze-thaw cycles.

Melissa and Lauren take elevations using a data collection unit and a prism pole

UMass Boston Historical Archaeology graduate students Melissa (left) and Lauren (right) use a data collection unit a prism pole to take elevations and lay out new units. These tools are used with a total station, operated by Dr. Schoenfelder (not pictured) to accurately map site coordinates.

Monday and Tuesday of week three also saw the arrival of Dr. John Schoenfelder and UMass Boston graduate student Lauren to the site. They worked with field school students to map further units near the Salisbury foundation, establish datum points at the Deb Newman site, and take some elevation measurements. This gave students attending the field school the opportunity to learn how to operate a total station and precisely map site coordinate

As this mapping was going on, Gary and Bryn finished their unit at the Augustus Salisbury site and moved over to the Deb Newman site to open the first unit there. This unit corresponds with some of the marked metal detection hits made by Brian in the first week. By Thursday they were joined by Liz and Melissa, while Alex, Andrew, Rick, and Tyler remain at the Salisbury site to finish their units.

While field work is generally supplemented by research and analysis after the season has officially ended, there are still times when you need to go home after a long day of excavating and consult a few books. After the discovery of a piece of pearlware or ironstone with a unique maker’s mark in a level suspected to be contemporaneous with the completion of the foundation, everyone ran to their phones to search the many online ceramic databases. When this proved to be too big a task for a quick in-the-field search it was decided that everyone would spend some time looking for the mark in online and print databases. Luckily the Fiske Center library is fully equipped for such a search. Three books containing examples of British and US pottery marksWhile the mark has yet to be identified, Dr. Mrozowski hopes that when it is it will give us a reliable TPQ (terminus post quem, or earliest possible date) for the Augustus Salisbury foundation’s completion. Perhaps the remaining units at the Salisbury site will provide more identifiable ceramic pieces from the same time period to aid in the TPQ determination.

The dual focus of this field school on both the Deb Newman and Augustus Salisbury sites provides an opportunity for comparative excavation and should prove very interesting in the coming weeks.



May 18, 2018
by John Steinberg

Twelve MA Theses in Historical Archaeology Defended in School Year 2017-2018


Kelton Sheridan talks about her MA thesis

Today at about 3 PM the Master of Arts Program in Historical Archaeology at UMass Boston will achieve a significant milestone:

More MA Theses were defended than new students accepted this year.

While the difference was only 2, it is important that this achievement be celebrated.   In addition to the 3 students that defended today:
Anya Gruber
Kelton Sheridan
Joe Trebilcock

And the 4 students who defended on Wednesday :
Sarah Johnson
Victoria Cacchione
Caitlin Connick
Leigh Koszarsky


We had 5 other students who defended earlier in the school year:
Caroline Gardiner
Alexandra Crowder
Ashby Sturgis
Jessica Hughston
Nadia Kline

This means 12 students defended during the 2017-18 school year. There are 10 graduate students in the 2017-18 matriculating Historical Archaeology class. There will always be some attenuation, thus having more students defend than enter will remain a very rare occurrence (as long as our program is thriving).  Our goal is that all of our students will finish their MA’s with an outstanding thesis and they will do it in a timely manner.

The Anthropology faculty and Fiske Center staff are constantly assessing the success of our MA program, not just by career path after leaving UMass Boston, but also looking at the time to degree.  The changes implemented over the last few years have probably made the MA even more rigorous.  At the same time, expectations and time tables have been more formally and clearly defined in the last few years.  That being said, most of the credit for this milestone goes to the hard-working students!

Just today there was an opinion piece in the New York Times by Ellen Ruppel Shell describing the financial consequences of not finishing an undergraduate degree.  While there are no statistics for Archaeology MAs, I suppose the costs of failing to complete the requirements are similar, though probably not as extreme. The success of our program depends on producing well-trained students who control the local archaeological sequences they are studying, deeply understand the unique and challenging archaeological methods they are using, and contribute to the theoretical problems in archaeology.   We will continue to work to put our students in a position to be successful.  Congratulations to all involved!

May 13, 2018
by John Steinberg

Spring MA thesis defenses for the Historical Archaeology program

Ground Penetrating Radar radargram from Chapter 4 of Joe’s Thesis

**Please join us for spring MA thesis defenses for the Historical Archaeology program.**

All defenses will be held in McCormack, 1-503. 

Wednesday, May 16, 2018

10 am, Sarah Johnson, “The True Spirit of Service”: Ceramics and Toys as Tools of Ideology at the Dorchester Industrial School for Girls.
11:30 am, Victoria Cacchione, “There are among the coloured people of this place remains of the Nantucket Indians”: Identity through Ceramics at the Boston-Higginbotham House.
1 pm, Caitlin Connick, An Analysis of Form and Function of Ceramic Rim Sherds from La 20,000, a 17th Century Estancia Outside Santa Fe, New Mexico.
2:30 pm, Leigh Koszarsky, Understanding Epidemic and Encampment: Yellow Fever and the Soldiers of Smallpox Bay, Bermuda. 

Friday, May 18, 2018

10 am, Anya Gruber, Palynological Investigations of 17th Century Agro-Pastoralism and Ecological Change at LA 20,000, New Mexico.
11:30 am, Kelton Sheridan, A Century of Ceramics: A Study of Household Practices on the Eastern Pequot Reservation. 
1 pm, Joe Trebilcock, Quantifying the Reliability of Ground Penetrating Radar at Archaeological Sites. 

June 15, 2015
by allisoncarlton001
1 Comment

Education and Excavation in Hassanamesit Woods

Dr. Mrozowski shows some of his field students how to map a feature.

Dr. Mrozowski shows some of his field students how to map a feature.

The small crew steadfastly completed their shovel-test pits and got to work on the larger unit excavations this summer in the Hassanamesit Woods. This year’s goal was to pinpoint the location of the late 18th/early 19th-century household of Deb Newman, who was a contemporary of Sarah Boston and the focal point of the project’s past excavation seasons. However, the shovel test-pits completed in the first few days of this year’s season were unable to gain any ground on that front. The field crew is currently focused on what is believed to be the nearby house site of Lewis Ellis, who was the son of a blacksmith with ties to Sarah Boston and Deb Newman.

Students excavate their units in Hassanamesit Woods.

Students excavate their units in Hassanamesit Woods.

Along the way, the students are getting a glimpse into the daily operations of an archaeological field excavation under the direction of Dr. Stephen Mrozowski. There are currently eight 2 x2  units being dug. The units have been placed according to historical maps and from reference to previous excavations in past summers. Throughout their progress, the students have uncovered an interesting material culture assemblage and some features that allude to an intriguing moment in the site’s history. The process has allowed students to understand the importance of historical documents as Dr. Mrozowski has conducted preliminary historical research to help make sense of the finds being recovered in the field.
The weather has been unusually cooler for this time of year, but this has allowed the crew to work hard and fast, and in the coming week this means expanding the search for Deb Newman and Lewis Ellis.

December 14, 2014
by John Steinberg

Geophysics at the Fowler Clark Farmstead in Mattapan

Using the CMD mini at Fowler-Clark

Using the CMD mini at Fowler-Clark

We are half way through a survey of the Fowler Clark Farmstead in Mattapan.  We were set back a little by the nor’easter last week, but will be out again finishing the GPR survey on Monday and Tuesday (December 15-16).

The geophysical work in on behalf of Historic Boston Inc., who would like to keep the pastoral setting of the farmstead. Today the 200-year-old farmstead sits on half an acre at Hosmer and Norfolk streets.  It is not known when the main farmhouse was built, but it appears on maps drawn between 1786 & 1806.  The barn is from about 1860.  You can learn more about this project on their blog which as a great 3D scan done by Feldman Land Surveyors.

We have some very preliminary results from the CMD.  The CMD is one of the instruments we were able to purchase with our recent NSF grant for work in Iceland from 2015-2017.  In 2013 we got a small grant to test these out in Iceland and like the unit very much, especially the temperature compensation.   That compensation algorithm turned out to be particularly important for the current November –December survey.

CMD 3 conductivity preliminary readings at Fowler-Clark

CMD 3 conductivity preliminary readings at Fowler-Clark

We surveyed with 25 cm transect intervals and fiducials mostly at 5 m.  This is the clipped conductivity 3 (largest dipole center distance – 1.18m)  readings.  The image mostly shows the distribution of sub-surface and near surface metal.

We will post more as we process it.


Skip to toolbar