The Lady and the Lake: Joe Cerniglia's Analysis of a Vial from the Colonial Village Site at Ritiati, Nikumaroro

The Lady and the Lake

by Joe Cerniglia

Editor's Preface: Another artifact analysis by one of TIGHAR's most redoubtable researchers. 

Please note the continuing need for data on pharmaceutical bottles from Australia and New Zealand. Finding a clear Antipodal match for the vial that Joe describes would be the simplest and surest way to disqualify it as an Ameliafact.  TFK

During the 2015 “In Search of Amelia Earhart” trip to Nikumaroro, organized by Betchart Expeditions, Kimberly Zimmerman spotted an amber medicine vial, resting on the ground among other glass vessels near the colonial village’s fallen co-operative store. The vial measures roughly 10 by 2 centimeters and has a flute-edged cap apparently made of aluminum. The cap has been punctured in the center as though with a sharp object. Both the cap and the base of the vial have machine-made markings. Since glass vessels in the village are assumed to be associated with the Nikumaroro colony, active from 1939 to 1963, we decided not to bring the vial back with us, but we took photographs. Since then, we have become interested in some of the vial’s attributes. We plan to retrieve this artifact when we return to the island in 2017.

Fig. 1: the amber vial

While the vial awaits recovery, the photographs we took provided some data to consider. We began by looking at the circular inscription on the top of the vial’s cap.

The Inscription on the Cap

We were struck by this inscription, which, on account of time and wear, is mostly illegible, save for a few consecutive letters: UCA | PHARM[1]

Fig. 2: The top of the cap

Since available literature supports the idea that amber vials were most commonly used for medicines,[2] it would seem safe to presume that the second word on the top of the cap is “Pharmacy.”

The word before “Pharmacy” is more problematic, since only 3 terminal letters, UCA, remain. Presumably, it represents the name of the pharmacy in which the vial was filled and capped.

While pondering this inscription, we noticed the intriguing coincidence between those three letters “UCA” and the fact that Earhart resided in Toluca Lake from early November 1934[3] until taking off from Oakland, California for what became her last flight on May 20, 1937. Her first home in Toluca Lake was a rented cottage on 10515 Valley Spring Lane. She and George Putnam moved into their last residence in the neighborhood on 10042 Valley Spring Lane on September 21, 1935.[4]

Additionally, we recalled the many other glass vessels found in various locales on the island that might have come from an American pharmacy[5], many of which are not easily placed in the context of a remote Pacific island inhabited by Tuvaluan and Gilbertese colonists (but which might be placed in the context of the U.S. Coast Guard Loran Station that operated on the southern tip of the island from 1944 to 1946). We have spent considerable time studying these other apparently pharmacy-based vessels.

There was no telling yet if the vial had come from an American pharmacy, and perhaps the Toluca Lake association was coincidental, but it would have been hasty to conclude this before setting out to discover what “UCA PHARM” may have meant, and how old the vial may have been.

Exploring Village Suppliers

The vial was found, again, on the site of the abandoned village colony. We know that the colony was supplied at various times by an Australian firm, On Chong & Company of Sydney.[6] The idea that the vial was supplied by an Australian pharmacy with the letters UCA at the end of a word in its business name was therefore a lead we definitely wished to pursue.

Unfortunately, we were unable to locate any historical directories of pharmacies in Australia. An exhaustive check of Australian sources of the vial was therefore not possible. However, we were able to locate a database of all present-day cities and towns on Earth, which included the names of locales whose names end in UCA, in Australia and elsewhere, thus suggesting pharmacies adopting those names.[7]

This world database showed there is only one town in Australia whose last three letters end in UCA: Echuca. An online search of present-day pharmacies in Echuca showed that there is an Echuca Amcal Pharmacy in business there. We received no reply from a letter we wrote to Echuca Amcal Pharmacy inquiring whether it had ever seen the amber vial, so our inquiries into possible Australian connections to the vial appear to have reached an impasse. We lack the data at present to investigate more thoroughly.

The relative proximity of Nikumaroro to New Zealand and the fact that the island colony was a British possession would argue for New Zealand and the United Kingdom as the next likely sources of origin for a vial found in the village. Again, however, we were unable to locate any historical directories of pharmacies in these locales. We did find, perhaps tellingly, that the world database showed there are no cities or towns in either of these countries whose names, or words within their names, end in the letters UCA.

Exploring American Leads

On the other hand, at least in the U.S., there was much more data at hand to collect and to analyze. A complete yearbook of U.S. drug stores in operation has been available since 1912 through an annual publication known as the Hayes Druggist Directory. From a detailed study of available Hayes Directories[8] and online pharmacy directories from 1922 to 2015, we created a database spreadsheet of all pharmacies in the U.S. whose business name, city, or street address has the letters UCA at the end of a word.[9] Each pharmacy that satisfied these criteria was placed at the head of its own column on this spreadsheet. The rows of the spreadsheet represent the calendar years 1922 to 2015, inclusive. If the pharmacy was in business for the year represented by a row, then a numeral was placed in the cell in which that calendar year (row) and that pharmacy (column) intersect. The numeral in relevant cells of the spreadsheet represents a “commercial strength” score of between 1 and 90, which is a rough comparator of the worth of inventory-on-hand of each pharmacy in each year it was in business, as computed by Hayes.[10]

Fig. 3: Hayes Directories on a table at the Chemical Heritage Foundation, Philadelphia Pennsylvania

Fig. 4: Hayes Directory key for commercial strength

By summing the commercial strength scores for each pharmacy in each column, a final score was computed for each pharmacy’s commercial strength for all the years it was in business from 1922 to 2015. By dividing each pharmacy’s final commercial strength score by the sum total of all the pharmacies’ scores, the relative proportional contribution of each pharmacy to the total commercial strength of the “UCA pharmacy market” was obtained as a percentage.

From this survey of directories we discovered:

·         Overall, U.S. pharmacies with UCA at the end of the business name, street address or city are rare. Only 26 were located in a 94-year period. There were six that were in business last year. Given that there were 60,276 pharmacies operating in the United States last year,[11] this means that roughly .01% (1 in 10,000) of the total U.S. pharmacies last year satisfied the criteria for inclusion in this analysis.

·         41% of the UCA pharmacies’ total commercial strength scores come from a single pharmacy in Amelia Earhart’s neighborhood, the Toluca Lake district of North Hollywood, California. This pharmacy was for most of its long history (1936-1991) known officially as the “Lakeside Pharmacy in Toluca Lake, North Hollywood,” but there was a brief period (1937-38) when it was listed in the San Fernando Valley City Directory simply as “Toluca Lake Pharmacy.”[12]

·         62% of the UCA pharmacies’ total commercial strength scores come from nine pharmacies, all with the word ‘Toluca’ in their name, within 3.1 miles of the Toluca Lake district, in North Hollywood, Studio City, and Burbank, California.

·         When the sample size is adjusted, such that only pharmacies with names ending in UCA are included in the database, discarding pharmacies occupying streets or cities whose names end in UCA, 88% of the remaining UCA pharmacies’ total commercial strength scores come from nine pharmacies within 3.1 miles of the Toluca Lake district of North Hollywood, Studio City, and Burbank, California.

We do not know whether Amelia Earhart was a customer of the Toluca Lake Pharmacy, much less whether she bought anything in vials there, but in three letters Earhart wrote to her mother before the world flight, she mentions a “vegetable concentrate” pharmacy elixir, telling of the benefits her husband, George Putnam, derived from taking it, and urging her mother, sister, niece and nephew to try some.[13] In one of the letters she states that a “Dr. Friend,”[14] would be willing to “make up a concentrate suitable for individual needs.” Vegetable and fruit concentrates often sold in vials, and still do today, as in the photo below.

Fig. 5: Modern fruit concentrate vials

Both of Earhart’s residences in Toluca Lake were between .7 and .8 miles of Toluca Lake Pharmacy. There was no pharmacy nearer by. The next nearest pharmacy to Earhart’s neighborhood in 1937 was W.D. Roberts Drugs, which was 1.4 miles distant.[15]

None of the research thus far, however, had attempted to answer the question of whether the vial was even old enough to have been brought to the island by Amelia Earhart, or whether it could be ruled out from having been brought by Earhart on account of its presumed age. To begin to answer this question, we turned our attention to the markings on the vial’s base.

Marks of Distinction?

Base markings can often identify a bottle’s manufacturer, the year in which a bottle was made, or a range of years in which a bottle could have been made.

On the base of the vial, there is a numeral 2 to the left of an unidentified central mark, and a numeral 0 to the right of it, along with a letter code, NT, centered below the mark.

Fig. 6: The base of the vial

The Society for Historical Archaeology’s Bottle Research Group[16] (a group that helps archaeologists and others identify bottles) and members of the Australian Antique Bottle Forum[17] examined these marks. Neither group was able to match the marks to any glass trademark they had seen or could reference.

The markings, however, share features with what the Owens-Illinois Glass Company used on its bottle bases from 1929 to 1954, as shown in the following photos:

Fig. 7: Toulouse (1971) illustration of OI mark with key explaining the meaning of the numerals[18]

Fig. 8: Compared vials

The green vial base on the left side of the above photo has the first Owens-Illinois trademark, "I" within a diamond over an oval, used from 1929 to 1954. The base of the amber vial found on Nikumaroro, on the right side of the above photo, has a letter "I" nested between brackets. Both vials have the same triadic arrangement of codes to the left, right, and beneath the mark. The Nikumaroro vial, however, lacks a central oval, shown in red, which, if it were present, would have heightened its resemblance to the Owens-Illinois base mark.

One way to explain the discrepancy relates to quality control problems. Glass factories used a lubricant called dope to prevent bottles from sticking to their molds. If a plant failed to clean the molds, whether from pressure to meet deadlines or simple oversight, the dope deteriorated and built up on the surface, so that letters or numbers on the finished bottle could be obscured or become very faint.[19] If degraded dope did build up on the mold that made the vial, it could have obscured the trademark of the Owens-Illinois Glass Company (or another company).

However, it goes without saying that facts, not hypotheses, are needed to identify a bottle mark conclusively. The only marks that can offer facts are those that are present, not those that may or may not be absent. Thus, although we hoped to learn the vial’s maker or its manufacture date, the markings on the base of the vial were simply too obscure to yield this information with certainty.

The Cap of the Vial: Datable Features?

We then turned our attention to the cap on the vial (excluding the inscription, which has already been discussed) in the hope that it would provide answers to our questions about when and where the vial was made. Three noteworthy features suggest the cap is of the roll-on variety, distinct from the far more common pre-threaded caps:

1)      The cap is conformed tightly to the vial such that it could not easily be removed.

Roll-on caps were, and are, “pressed in the capping machine by rotating rollers and shaped to conform with the individual bottle’s threads.”[20] While one cannot discount the possibility the cap is misshapen from repeated use, or originated from a container other than the vial itself, poor fit was a common complaint of caps in general in the early 1930s.[21] R-O caps specifically were known to stick because defects in the threads of the finish could cause the cap to be “drawn into the defect, locking the cap on and making it very difficult for the consumer to remove.”[22]

2)      The cap is made of heavy gauge metal, probably aluminum, and is not rusted. It has no outwardly rolled distal wire edge, as would be common on ordinary pre-threaded screw caps.

The Aluminum Seal Company (Alseco), a subsidiary[23] of the Aluminum Company of America (Alcoa) of New Kensington, Pennsylvania, first introduced roll-on caps, also known as R-O, in 1924. R-O caps were “initially used” in the early years after 1924 “for prescription drugs and later in some vapor-vacuum sealed caps for foods.”[24] While not all aluminum caps are R-O caps, it would appear that all R-O caps are aluminum, due to aluminum’s excellent malleability.[25] They can be found today on liquor bottles and some condiments, such as olive oil.[26]

3)      The cap has a rectangular tab-like indentation a few millimeters wide on the lower part of its skirt, in which the metal appears to have been broken off from the cap.

The rectangular indentation, visible in the lower right corner of the cap on the amber vial in the photo below, could be a remnant of a tamper-proof seal patented in 1933 by Alcoa specifically for R-O closures.[27] The seal was pressed around the base of the cap during the production process of crimping the cap with rollers. One removed the seal by pulling up on a tab. Once removed, a portion of the depending metal skirt could be deformed or torn away.

Fig. 9: Amber vial and 1936 Seagram's advertisement for pilfer-proof R-O cap

Fig. 10: Amber vial and close-up of Seagram's advertisement

In summary, then, we have:

·         The poorly fitting aluminum cap, probably R-O, introduced in 1924 by Alseco, and

·         The rectangular tab of missing cap material, possibly representing a pilfer-proof seal patented in 1933 by Alcoa, parent of Alseco.

If we are interpreting these features correctly, they could indicate the cap was made between 1924 and the mid-1930s.

The cap could also have been made at a later time, however, since the use of R-O seals on drug containers continued long past the 1930s. Although aluminum closures “seem to have been banned altogether” during World War II,[28] and despite aluminum shortages during the Korean War (1950-53),[29] R-O caps on drug bottles were marketed in the 1950s and 1960s, and “might yet be in use in Central America.”[30] These later R-O caps are described in trade literature as having a “locking band attached to the cap by small metal bridges… (The locking band) remains on the bottle after the cap is removed.”[31] One can easily locate this type of R-O cap today on condiments sold in supermarkets.

While the Nikumaroro vial does not have a persisting locking band, such a band could have been removed from the cap during its history.

In general, however, R-O aluminum closures on medicine containers appear to have been very rare. We surveyed hundreds of medicine vials and pill bottles at auction on eBay and Etsy.com. With one exception, all the caps on these containers, dated from 1932 to 2003, appeared to be pre-threaded, not roll-on. A very small number appeared to be aluminum, but the vast majority appeared to be tinplate or Bakelite. We obtained a sample of these containers at auction. All the caps on the bottles from this sample were easily removed.[32]

The only medicine bottle we found, in addition to the vial itself, that appears to have an aluminum R-O cap was a bottle of topical cocaine ointment, pictured below. The bottle has the locking band as described in postwar trade journals, and a paper label dated February 2003.

Fig. 11: Cocaine hydrochloride bottle with R-O cap

We have been seeking but lack data concerning when R-O closures were used on drug containers in countries other than the United States. It could be that the absence of data indicates this type of cap was seldom or never used on drug containers elsewhere in the world, but we cannot say for sure.

A Knurly Problem

Experts from the Bottle Research Group have identified one feature of the vial, however, that causes them to say it is definitely not from the 1930s and thus could not possibly be Earhart’s. They say the little tooth-like marks on the perimeter of the base are datable to the 1950s or later.

Fig. 12: Perimeter knurling on base of the vial

The tooth-like marks are an example of what is known as resting point knurling, defined as marks impressed by machine to the bottle mold, which in turn create marks on the perimeter (the resting point) of a slightly concave bottle base. According to Russ Hoenig, a former Owens-Illinois shop foreman, these marks offered protection to the glass against cracking. These and other like marks “became a necessity as bottles were light-weighted with thinner glass. The cold checks (cracks) present on all bottles formed prior to annealing now penetrated proportionately deeper through the glass wall thickness on lightweight flow-manufactured bottles, which caused breakage.”[33]

The earliest documented date assigned to resting point knurling of the type exhibited on the Nikumaroro medicine vial is 1958. Coincidentally, that date is based on another Nikumaroro glass vessel, a green beverage container upturned as a grave marker on the western shore of the island, a few hundred meters north of the landing channel (editor's note: the Ritiati Bottle Grave, recorded in 1997 and again in 2015, now heavily eroded). That bottle has a date code of “58” to the right of the second Owens-Illinois trademark, which first appeared on bottles in 1954.

Fig. 13: Green beverage bottle in grave on western shore of Nikumaroro

Base markings used to prevent cracking could, as in the case of the amber vial, surround the perimeter of a bottle base, but much more often they covered the entire base. Where such marks were impressed on the mold’s base plate by machine, they are called full base knurling. Where such marks were impressed on the mold’s base plate by hand, they are called full base stippling. According to the Bottle Research Group, full base knurling by machine first appeared in 1945, and full base stippling by hand first appeared in general use in 1940.[34] [35]

However, the timeline constructed by the Bottle Research Group may not be perfect. There is evidence that full base stippling by hand appeared on select drug containers as early as 1932.

Antidotes

We have located two identically designed “Abbott Lab” pill bottles in amber with full base hand stippling, identical base markings (including the same plant number and year code), and identical prescription wraparound labels.

Fig. 14: Twin pill bottles

Fig. 15: The base of one of the Abbott Lab pill bottles

The label on one of the bottles has been sunned to the point that most of the printing is no longer legible. The label on the other is legible, but the typewritten year of its prescription date is slightly faded. To help determine exactly what the year on the label of this bottle is, we consulted a board-certified document examiner. Emily Will, D-BFDE (http://qdewill.com), stated her professional opinion, based on rigorously certified techniques, that the year of the label’s prescription date is 1934.[36]

On the base of the bottle with the dated label, the embossed year-of-manufacture code, situated to the right of the first Owens-Illinois trademark (used from 1929 to 1954), is a 2. This represents a year ending in the numeral 2. We presume that the bottle was manufactured in the year of the prescription date on the label, or in a prior year, such that the bottle’s year of manufacture must be earlier than or equal to 1934. The only calendar year between 1929 and 1954 that ends in 2 and is earlier than or equal to 1934 is 1932. Therefore, it is reasonable to presume this bottle with full base stippling and a dated paper label was manufactured in 1932.

We have located another amber medicine bottle, also with full base stippling. This bottle’s design is patented, as indicated by a patent number, 94824, stamped on the base. The U.S. Patent Office approved the patent in 1935.[37] The plant number, situated on the base to the left of the first Owens-Illinois mark (1929-1954), is Owens-Illinois plant number 4, Clarksburg, West Virginia, which went idle in 1944.[38] The year-of-manufacture code, situated on the base to the right of the mark, is a 9, which signifies a year ending in the numeral 9. Based on the Clarksburg plant’s years of operation, the bottle cannot be dated to 1949. Based on the patent date, it cannot be dated to 1929. Therefore, the bottle is dated with certainty to 1939, a year prior to the earliest previously documented example of full base stippling.

Fig. 16: Base of 1939 amber medicine bottle

While we have yet seen no glass container earlier than 1958 with the same resting point knurling as found on the Nikumaroro vial, the pill bottle and the patented amber bottle demonstrate that no dating system is perfect. Careful research may occasionally uncover exceptions to any archaeological dating rule. The Bottle Research Group looked closely at beer and soda bottles for relevant examples when it constructed its timeline, but it appears that it did not inspect medicine containers.[39]

An Amelia-fact?

Since full base stippling was first used on some medicine containers a few years before beer and soda, it may be that perimeter knurling followed a like pattern, first appearing on select medicines in the 1930s and then later on beverages in the 1940s once it was needed for the lighter weight bottles. We do not know whether these speculations are true, but they are logical. If they are true, and if the vial really is from the U.S., then Amelia Earhart just may have brought the vial from Toluca Lake Pharmacy to the island. At least the possibility should not be ruled out. But we have other caveats, such as:

·       If the vial was Earhart’s, it likely took a circuitous route to the village, which is “not where we are likely to find Earhart-related objects in their original contexts.”[40] Instead, it was likely brought there from somewhere else.

·       We do not know the full list of pharmacies in the world whose name, street, or city end in UCA, but, as has been stated, we do know that the colony was supplied by an Australian firm, which could easily be a source of pharmaceutical vials.

·       We know that at least five Coast Guardsmen at the Loran Station on Nikumaroro between 1944 and 1946 lived before the war in U.S. cities or towns that were within driving distance (1 to 100 miles) of several towns with pharmacies in the database of UCA pharmacies.[41]

Conclusions and Questions

Some of the Nikumaroro artifacts fascinate because they may be of a certain time, the 1930s, or of a certain person, a castaway. The amber vial fascinates because, although it is not proven to be, it just might be of a certain place, a little village in California that once counted Amelia Earhart among its own – along with a friendly drug store, named Toluca.[42]

Since the island, however, had its own village, and a history of colonists, Coast Guardsmen, explorers, researchers and others, who doubtless used medicines, there exist other equally or more equally plausible ideas as to the amber vial’s provenance. We have ideas, but few definitive answers.

Nevertheless, there are some questions that might be answerable, definitively we hope, through further research, to which anyone is welcome to contribute:

1. Have you seen the markings found on the bottom of the vial on other glass vessels? Can you say what they mean?

2. Can you locate any of the following at auction, antique shops or archaeological sites?

·         R-O closures on medicine containers

·         Medicine containers with resting point knurling

·         Base-stippled medicine containers whose manufacture date is from the 1930s

·         Toluca Lake Pharmacy vials

3. Do you know of other pharmacies outside the U.S., in Australia or elsewhere, with names ending in “UCA”? Do you know of any we missed inside the U.S.?

We also have some questions that might be answerable by those intrepid researchers fortunate enough to be on the next cruise to Nikumaroro in 2017:

A. What types of medicine containers (vials, pill bottles, etc.) can be found near the village dispensary and elsewhere in the village?

B. Can the amber vial be located again, recorded for its context and associations, and retrieved for further analysis? If so, what tests might bring back the most relevant data?

Research will continue on these questions and, very likely and in keeping with the nature of a mystery, many more that we have not yet thought to ask.

Selected Bibliography

Backus, Jean L. Letters From Amelia: An Intimate Portrait of Amelia Earhart. Boston: Beacon Press, 1982.

Bender, Nathan E. Historic Bottle and Jar Closures. Walnut Creek, CA: Left Coast Press, 2016.

Giles, Geoff A. Handbook of Beverage Packaging. Boca Raton: Taylor & Francis, 1999.

Griffenhagen, George and Bogard, Mary. History of Drug Containers and Their Labels. Madison, WI: American Institute of the History of Pharmacy, 1999.

Hanlon, Joseph F., Kelsey, Robert J., Forcinio, Hallie. Handbook of Package Engineering, Third Edition. Boca Raton: CRC Press, 1998.

Lief, Alfred. A Close-Up of Closures: History and Progress. New York: Glass Container Manufacturers Institute, 1965.

Osborn, Albert S. Questioned Documents, Second Edition. Albany: Boyd Printing, 1950.

Peterson, Arthur G. 400 Trademarks On Glass. DeBary, Florida: Arthur G.Peterson, 1968.

Sonnedecker, Glenn; Cowen, David L.; Higby, Gregory J., eds. Drugstore Memories: American Pharmacists Recall Life Behind the Counter, 1824-1933. Madison, WI: American Institute of the History of Pharmacy, 2002.

Toulouse, Julian Harrison. Bottle Makers and Their Marks. Caldwell, New Jersey: The Blackburn Press, 1971.

Winters, Kathleen C. Amelia Earhart: The Turbulent Life of an American Icon. New York: Macmillan, 2010.

For information on the Nikumaroro hypothesis concerning the 1937 disappearance of Amelia Earhart, see:

Gillespie, Richard. Finding Amelia. Annapolis: Naval Institute Press, 2006.

King, Thomas F., Jacobson, Randall S., Burns, Karen R., Spading, Kenton. Amelia Earhart’s Shoes. Lanham, MD: Altamira Press, 2004 (updated edition).

King, Thomas F., Thirteen Bones. Indianapolis: Dog Ear Publishing, 2009.

---

[1] We interpret the vertical line between “UCA” and “PHARM” as an interpunct, a word separator.

[2] Griffenhagen, George and Bogard, Mary. History of Drug Containers and Their Labels. Madison, WI: American Institute of the History of Pharmacy, 1999, pp. 35-36.

[3] Winters, Kathleen C. Amelia Earhart: The Turbulent Life of an American Icon. New York: Macmillan, 2010, pp. 151-152.

[4] Papers of Amy Otis Earhart, Arthur and Elizabeth Schlesinger Library, Radcliffe Institute for Advanced Study, Harvard University, Cambridge, Massachusetts. Call No. MC 398, M-129, Reel #3.

We are indebted to Sarah Hutcheon, Research Librarian at the Radcliffe Institute for Advanced Study, for her assistance.

[5] For a list of various containers of suspected pharmaceutical origin, see Dr. Thomas King’s blog at http://ameliaearhartarchaeology.blogspot.com/2016/06/2017-nikumaroro-draft-research-plan-i.html

[6] See http://tighar.org/Projects/Earhart/Archives/Documents/Tarawa_Archives/1939_Co-op_Store/1939store_3.pdf for details on the Gardner colony’s relationship with On Chong & Company.

[7] See https://www.maxmind.com/en/free-world-cities-database

[8] On November 23 and 24, 2015, The Chemical Heritage Foundation in Philadelphia, Pennsylvania generously opened up its collection of Hayes Directories for research purposes. We are indebted to librarian Ashley Augustyniak for her assistance.

[9] The pharmacy database may be viewed at https://www.dropbox.com/s/h34l05afr89gbbo/Pharmacy%20Database%20and%20Loran%2092%20Census.xls?dl=0

[10] For pharmacies in business after 1997, we were unable to access the relevant Hayes Directories. Instead, we used pharmacy directories available online from insurers. For these pharmacies, we estimated the commercial strength scores based on size of the market. For those years in the 20^th century for which no Hayes Directory was available, we filled in missing data by using information from Hayes Directories from years prior to and later than the gap, when it was logical to do so.

[11] www.hayesdir.com/aboutus.html

[12] Yellow pages: San Fernando Valley Directory. 1937-38. Long Beach, CA: A to Z Directory Publishers.
See the Directory listing for Toluca Lake Pharmacy at https://www.dropbox.com/s/rm8b6z4gva89obf/Photo%20Oct%2031%2C%201%2028%2042%20PM.png?dl=0. Note that this establishment is also listed elsewhere in the directory as “Lakeside Pharmacy.”

[13] Earhart, Amelia. Letters to Amy Otis Earhart. 5 July 1935; 25 Feb 1936; 20 Mar 1936. MS. Papers of Amy Otis Earhart, Arthur and Elizabeth Schlesinger Library, Radcliffe Institute for Advanced Study, Harvard University, Cambridge, Massachusetts.

[14] The chief pharmacist, owner and founder of Toluca Lake Pharmacy was Robert G. Eyth. Eyth is an Anglo-Saxon word that is loosely translated as “friendly.” The 1937-38 San Fernando Valley telephone directory lists no doctor by the name of “Friend.” In Letters from Amelia, a compilation of letters written by Earhart, editor Jean Backus chose to redact the name of Dr. Friend, instead referencing him or her as “the doctor.” She presumably did this to mask the doctor’s identity (for reasons unknown).

[15] Yellow pages: San Fernando Valley Directory. 1937-38. Long Beach, CA: A to Z Directory Publishers. Distances computed by mapquest.com.

[16] The Bottle Research Group consists of:

Bill Lockhart, Professor, New Mexico State University

Dr. Peter Schulz, Senior Archaeologist (retired), California State Parks, Sacramento, California

Carol Serr, Archaeologist, Laguna Mountain Environmental, Inc., San Diego, CA

Bernard Schriever, Archaeologist, Laguna Mountain Environmental Inc., San Diego, California

David Whitten, collector and glassmaking historian, Clarksville, Indiana

[17] www.ozbottleforum.com

[18] Toulouse, Julian Harrison. Bottle Makers and Their Marks. Caldwell, New Jersey: The Blackburn Press, 1971, p. 403. 

[19] Lockhart, Bill. “Re: Photo from Bottle Grave Site.” Message to Joe Cerniglia and Thomas King. 19 Dec 2015. E-mail.

[20] Lief, Alfred. A Close-Up of Closures: History and Progress. New York: Glass Container Manufacturers Institute, 1965, p. 29.

[21] Ibid., p. 33.

[22] Giles, p. 238.

[23] https://en.m.wikipedia.org/wiki/Alcoa

[24] Hanlon, Joseph F., Kelsey, Robert J., Forcinio, Hallie. Handbook of Package Engineering, Third Edition. Boca Raton: CRC Press, 1998, p. 436.

[25] Giles, Geoff A. Handbook of Beverage Packaging. Boca Raton: Taylor & Francis, 1999, p. 238.

[26] For a demonstration of the process of creating an R-O cap, see the video at www.fowlerproducts.com/videos/roll-on/

[27] To view the patent, see https://patentimages.storage.googleapis.com/pdfs/US1895884.pdf

[28] Bender, Nathan E. Historic Bottle and Jar Closures. Walnut Creek, CA: Left Coast Press, 2016, p. 90.

[29] U.S. Congress, Office of Technology Assessment, Technical Options for Conservation of Metals: Case Studies of Selected Metals and Products, OTA-ITC M97 (Washington, DC: U.S. Government Printing Office, September 1979), p. 118.

[30] Bender, pp. 84-86.

[31] Modern Packaging Encyclopedia, 1951 edition., Packaging Catalog Corporation, 1951, p. 456.

[32] For a database of medicine containers, see https://www.dropbox.com/s/yi66g54yi1g5wdu/medicine%20container%20survey.xls?dl=0

[33] Lockhart, Bill. “Re: The Boston Pill Bottle.” Message to Joe Cerniglia and Thomas King. 27 Jan 2016. E-mail.

[34] See https://sha.org/bottle/bases.htm

[35] Lockhart, Bill. “Re: The Boston Pill Bottle.” Message to Joe Cerniglia and Thomas King. 7 Dec 2015. E-mail.

[36] For Emily Will’s full forensic report, see https://www.dropbox.com/sh/zqp92wn8wb17xms/AABC2IouXb_YFz4sD6_wHGDva?dl=0. For a biographical timeline (source: Ancestry.com) of persons whose names appear on the label, see https://www.dropbox.com/s/d7fhc702fksm3ge/Photo%20Jul%2005%2C%2011%2005%2046%20AM.png?dl=0. The timeline demonstrates the label is internally consistent with the prescription date of 1934.

[37] http://www.google.com/patents/USD94824 and also https://www.sha.org/bottle/pdffiles/USD94824.pdf

[38] https://sha.org/bottle/pdffiles/O-IPlantList.pdf

[39] The archaeological site from which examples of bottles with stippling and knurling were obtained was the back lot of a former beer distributor in El Paso, Texas. Lockhart, Bill. “Re: The Boston Pill Bottle.” Message to Joe Cerniglia and Thomas King. 7 Feb 2016. E-mail.

[40] Thomas F. King, http://ameliaearhartarchaeology.blogspot.com/2016/06/2017-nikumaroro-draft-research-plan-i.html

[41] Using Ancestry.com, we compiled a census of all 71 men who inhabited Loran Unit 92 on Nikumaroro, to try to discover each man’s residence just prior to World War II. See the census at the second tab of the Excel document at https://www.dropbox.com/s/h34l05afr89gbbo/Pharmacy%20Database%20and%20Loran%2092%20Census.xls?dl=0. The personnel roster was sourced from www.loran-history.info/roster.aspx.

[42] For a circa 1959 tour of shops along Toluca Lake's main road, Riverside Drive, which includes a Toluca Pharmacy at 0:55, see http://youtu.be/3Rk1r6M0G8I