The Stories of the Maps

The Spotting Map -- Credits

The online spotting map (the one with dots showing the locations of the buildings selected in the study) is based on two paper maps from the collection of the Lowell Historical Society. They are available in A Portfolio of Historical Maps of Lowell 1821-1914, published by the Lowell Historical Society, P.O. Box 1826, Lowell, Massachusetts. The two maps used are:

  • Present day Lowell showing dates of annexation -- 1914. [Originally published by Sampson & Murdock]
  • Ethnic Districts - 1912 - from "The Record of a City" - by George F. Kenngott

The portfolio (containing six maps, right) is available for purchase at the Visitor Center of the Lowell National Historical Park for $8 (as of 2005).


The map portfolio is from the collection of the Lowell Historical Society.
By the way, we are using a map dating from 1914 since most of the stories are closer to that date than they are to the present and that map closely approximates the city for a couple decades either way. The map doesn't match the current city of Lowell for two reasons. First, the city has changed significantly since then -- buildings have been built and torn down, urban renewal destroyed neighborhoods, entire mill complexes disappeared, a freeway displaced many neighborhoods, streets were relocated -- all the things that happen in a city over a hundred years. It's tempting to try to overlay a modern map on the city to show the difference. However, the civil engineering surveying methods of the time don't match the exquisite detail we currently see in aerial maps and GPS routing. The older maps just don't line up -- parts are significant skewed and the exercise is very frustrating.

Digitizing the 1914 map -- geek talk

Section of Lowell map overprinted with annexation dates

The base map (about 14"x10.5") was scanned at 600 dpi (in June, 2003. The annexation information that had been printed over many areas of the map was erased and the underlying image repaired as much as possible (at the cost of sore hands and eyes for the map maker). Go ahead, click the map at the left and see how the erasing was done. Ooooh! Isn't that neat? Click again. Ahhhh...

The ethnic districts were colored in by hand (okay, okay, it was by mouse). The result was a file 8544 pixels wide x 6264 pixels high x 24 bits of information per pixel. In Photoshop(R) file format it was about 93MB. Having gone this far, it seemed reasonable (sure it did, sheesh!) to add colors for the rivers, canals, trolley lines, and railroads, as well as labels for the ethnic communities and a delineation of the 1840 Irish areas (as found in the NPS booklet for the park). With great regret, no pear tree was added. This resulted in a file of 108MB. Those experienced with image processing might guess this proved just a mite cumbersome. The file was rescaled to 5000x3624 pixels and then rotated 20 so that north was (almost) up like all other maps in the world. After rotation it was re-cropped to 5000x3500. That brought the size down to down to a mere 47MB. Since it was still painful to work with, this map's greatest virtue was producing the final justification for replacing the map worker's four-year-old computer with a new, whiz-bang, golly-gee, fast one. The final image touchups, addition of the E&E project's research locations at the correct street locations (the red and purple dots), and production of the web page versions were now possible.

Get a better map! Amaze your friends!

Especially for those with broadband connections, get the all-singing, all-dancing version: the entire city of Lowell in 1914 -- not just the center section. And more, it has the location dots and the colored ethnic areas you've learned to love and appreciate on previous maps. But wait, there's more! Trolley and railroad lines are colored in! This super map has the same resolution as the size on the site-- it just covers more area (5000x3500 pixels instead of 2800x2000). You can download our biggest, best quality map for the low, low, low price of 1,001,875 bytes (230 seconds at 56K, 10 seconds with broadband)! (Truth in advertising: The dots aren't hyperlinks on this version.) Alternately, get this best quality without decorations (no coloring, no dots, just plain old black and white) at just 1,003,976 bytes. (Why is the plain map bigger than the fancy one? We have no idea. They were written with the same compression factor.)

What? You want to download the 8544x5254 map? Get real! Even the mapwebster finds its size unmanageable.

The 1924 Lowell Atlas

Title page of the "Richard's Standard Atlas of The City of Lowell, 1924"

The Center for Lowell History has an online "Digital Map Collection of Lowell." Among the maps (as of July 2008) are atlases from three years, 1879, 1896, and 1924. These are invaluable resources for tracking where people lived and worked and provide valuable insights into relationships.

The 1924 maps were provided by the Lowell National Historical Park. They were originally printed in large folios about 2 by 3 feet in size. For 1924, there are are 33 large map sheets.  One is a "key" map of the whole city with an attempt to name all the streets so you can use this map as an index into the others. To show the whole city, there was some liberty taken in the placement of streets so the large map isn't very definitive.

The remaining 32 maps each show a small section of the city, usually quite accurate, containing details of buildings and addresses.  The publishers broke out the small sections for convenience of publication. That is, they arranged the individual maps on the sheets to provide the greatest density and least number of sheets.  Unfortunately, this is very inconvenient for map readers going from one sheet to another because north is very seldom in the same direction on adjacent maps.

After using these for a while and getting frustrated, I edited the maps, rotating them so that north was always up, making it a bit easier to navigate.  I created a series of pages making it easy to move from the index map to the individual section maps and to move among the section maps.  Since the maps as published at the CLH site are quite large (on the order of 2000 x 3000 pixels) it became necessary to figure out how to pan and zoom along with the map to map navigation.  Finally, it got to the point that I just dedicated about six weeks on the maps to make future research more productive.  I recently polished up the interfaces to these homegrown hacks and am now publishing them for all the other Lowell map freaks (both of you.)  Have fun looking at the 1924 Atlas! (July 30, 2008)

(Along the way, I also applied the panning and zooming techniques to the spotting map, making that easier to use instead of the two different sized maps previously.)

Now that I know how to do it, the 1879 and 1896 maps will go faster, if I ever get around to them, which would definitely be useful.  Just give me another six weeks of free time....

The 1896 Lowell Atlas

Right, just six weeks of free time should do for the next set of maps.  Oops.  It took almost four months.

The 1896 maps were provided by the Lowell Historical Society and are available in original form at the Center for Lowell History site. There is a key map, 31 local maps, and a street index. It is organized identically to the 1924 Atlas mentioned above.

The 1896 Atlas was more complex to adapt since it was printed in vibrant colors whose uneven fading made legibility more of a problem.  I foolishly tried to restore more of the 1896 colors and tried to make the background brightness more equal than I did for the 1924 maps. I learned a lot as I did many maps. The later ones are better than the earlier ones but I didn't always go back to improve the earlier ones since I was getting thoroughly tired of the whole project.

A few notes: The maps here are not exactly as printed in the original atlas -- they're better. There were several plates with insets that were actually extensions to the map (they would have run off the page). Since we have to pan to see all of the map in a browser anyway, I took those insets and positioned them at the correct place on the map, extending the map in that direction. (I did this also with the 1924 maps, but there were only a couple of those.)  Also, Plate 20 was printed with an incorrect directional arrow for North. I compared the streets in Plate 20 with the streets of adjacent maps and made an adjustment for Plate 20 that is consistent with North on adjacent maps. (Strangely, North is not always the same direction on adjacent maps or the key map. Why didn't they check their GPS readings in 1896?)

The rotation of each map to the North and the color adjustments may make the "corrected" map more difficult to read than the original although, of course, I tried not to let this happen (hence, four months of work).  Just in case, I have provided a link on each map page to the original map on the Center for Lowell History website so you can check to see if the original is any better. I added this link to the 1924 Atlas pages as well.

So, finally, the 1896 Atlas is now available. (April 6, 2009)

I'm giving no estimates for the 1879 Atlas. It may never get done.


The 1879 Lowell Atlas

The 1879 Atlas is now available. (January 10, 2010) These 1879 maps were originally provided by the Lowell Historical Society and are available in original form at the Center for Lowell History site. There is a "base" map and 24 detailed maps, but no street index.

I learned a few tricks doing these maps so they now all look quite similar to each other in gray levels. As with the other maps, a few maps had insets that I repositioned, making the maps wider or taller than the originals.

What is North?

Some of the maps' north arrows were clearly in error so I did a little digging. I compared all the arrows with true north as defined by the City of Lowell Geographical Information System (GIS) and by Google Maps which were, thankfully, in agreement. Of course, the old maps were often at variance with the precise modern versions -- streets were at incorrect angles and distances were inconsistent. Even the town boundaries, which one would think should be very accurate (after all, it affects taxes!), were sometimes wrong. However, it was possible to make good comparisons (probably less than a degree error) on all but two maps. The arrows on only four maps were reasonably accurate, within a degree of true north. Two maps had arrows that were east by 1.5 and 1.8.  The remaining nineteen (of twenty-five) had arrows west of north by anywhere from 1.5 to 17.9, with the base map itself off by 13.4.

One of the maps gives a possible clue about the error. Plate U has two arrows, one labeled True Mer, within 0.5 of north -- essentially correct, given printing technology of the time. The other arrow points west of north and is labeled Magnetic Mer / Variation 1130", suspiciously close the to average west error on the maps of 10.2. That doesn't explain, however, why the maps are off by widely different amounts.

In any case, I oriented all maps so that true north (according to GIS) is up and added a comment near each arrow about its accuracy. It's much easier to navigate from map to map when north is in the same direction.

I hope to repair the orientation of the maps in the other atlases sometime soon. (Wanna buy a bridge?)



No mills were destroyed in the production of these maps. No boarding houses were occupied or restored, no textiles woven, and no tours cancelled or conducted. All pixels are verified volunteers.
This page does not represent the official position of the University of Lowell or the National Park Service. Merely allowing it to appear is the price paid to the volunteer map maker, who just had to tell the geeky part of this project story.