As is well known, vast numbers of Jews left Eastern Europe around the turn of the 20th century, destined primarily for the United States—but also for South Africa, Palestine, Argentina, and elsewhere. The ships that carried them also brought passenger manifests, documents that presumably listed all passengers and provided considerable information about them—documents prized by genealogists for the data they provide.

“The name was changed at Ellis Island.” That hoary, but cherished, urban myth popped up at a recent talk I gave to an audience composed primarily of non-genealogists with an interest in the topic. When told that this likely did not happen, the audience reacted with disbelief and began to asked detailed questions about the creation of U.S. immigrant passenger arrival records at the turn of the 20th century. Who decided what was needed? How was the information provided, and who recorded the data? Did the manifests list passengers other than those in steerage?

The questions asked prompted me to re-examine exactly what I do know (or did not know) about the entire immigrant arrival procedure into the United States and elsewhere at the turn of the 20th century. I turned to experts Nick Evans, Saul Issroff, and Marian Smith for enlightenment.

Immigration to South Africa

Sallyann Sack-Pikus: How exactly did the process work in the case of South Africa? Did all Litvak (Lithuanian) immigrants (or nearly all) go through London?

Saul Issroff and Nick Evans: The majority went through London, but after 1896, some went via Hamburg and Southampton and fewer via Bremen. A very few seem to have come via Rotterdam or Le Havre.

Sack: Did they all travel on the Union Castle Line or some other British steamship company?

Issroff and Evans: Usually it was the Union or the Castle lines—and after the lines merged, the Union Castle Line. A small number traveled via other continental lines, and some went on vessels destined for Australia or New Zealand that stopped in Cape Town.

Sack: Did all or most immigrants purchase tickets for travel to South Africa back in Lithuania from British ticket agents?

Issroff and Evans: Tickets could be purchased in New York, London, Paris, Cape Town, Johannesburg, and many small Baltic towns. From there, details probably were telegraphed to the points of departure. In other words, they could be purchased in any locale that had a telegraph office.

Sack: Did the ticket agents supply the names and other identifying information to the steamship companies? Did the immigrants have to present any written documents anywhere along the line?

Issroff and Evans: Certainly it appears that at Libau some documents had to be shown—and also at Southampton and London. As we understand it, the purser prepared the manifests from information supplied by the ticket agents. Many changes are noted on the initial lists. For example, people who did not board the ship are crossed out.

Sack: Did the South African authorities require any documentation to enter the country in the years of major Jewish immigration there?

    Issroff and Evans: The only material we have seen relates to the Boer War period (Winter 1900 to Spring 1902) during which the British army issued some sort of landing permits.

    Sack: My Shulkin relative from Lithuania became Wolkin in South Africa, according to family lore, “because a South African immigration official misread the Cyrillic letter that makes the ‘sh’ sound (and actually is a Hebrew shin) for the letter W.” Where exactly in the process would this have occurred?

    Issroff and Evans: This is the million-dollar question. No one will ever know. From evidence in Britain and the British Empire, no one ever had their name changed based on passenger lists.

Immigration to the United States

    Sallyann Sack-Pikus: What papers did a prospective immigrant have to show when he bought passage to the U.S from Europe? We Jews have heard a couple of seemingly contradictory things. For example, one of my grandfathers told me that he adopted the name Steinsnider (although his real name was Dubner), because he was “running away from the Russian army and bought papers from a guy named Steinsnider.” To me that suggests that he had to (or thought he had to) show some sort of papers when he purchased passage—and that the clerk who compiled the manifest used those papers to make the entry for my grandfather. In other words, the manifest information was taken from written (not oral) material. On the other hand, we have heard numerous stories about Jewish men “sneaking” over the border to get out of Russia, presumably not showing papers to anyone on exit.

Retired Hamburg, Germany, city archivist, Jurgen Sielemann once gave a talk at a conference in which he described how Hamburg ticket agents would go into Kovna guberniya, sell tickets, and arrange for the men to sneak over the border. That scenario suggests to me that they did not necessarily have to show documents and that some/many/most manifests may have been compiled by clerks who listened to information given orally to them prior to passage by the emigrants. And then there was the question of woman and children who came later and did not sneak across the border. Did they need documents in order to leave?

Marian Smith: The answers to your questions depend on when, where, and to whom.

Prior to 1918, the United States did not require an immigrant to show any document to U.S. authorities at the port of arrival other than the manifest submitted by the ship captain (more on that below). That said, many immigrants were asked by the immigrant inspector to produce a letter mailed within 90 days from the friend or relative at their destination, just to confirm that they were expected.

Similarly, the steamship companies required no documents. Most tickets were purchased from agents and not directly from the company. The agents obviously would have preferred to compile their information from written documents so, in any case, may have written part or all of the manifest information by copying from another document. It was the ticket agent who forwarded the information to the steamship company whose clerks then used the agent’s records to compile the passenger list.

In the case of pre‑paid tickets, the immigrant’s information came from his or her husband/brother/cousin in Baltimore (or some other U.S. city) rather from him or herself. The immigrant just collected the ticket before departing.

Foreign exit document requirements varied widely by country. For example, Russia required male emigrants to obtain an exit permit before they could leave; to get one, they had to have completed their required military service. The Russian government also had an understanding (if not a formal agreement) with the railroads and shipping companies not to sell tickets to anyone who could not produce an exit permit showing military service. Hence, the common story of young men who bought or otherwise obtained someone else’s military papers to get an exit permit. Once they did so, they had to buy the ticket under that assumed name. Once that was done, the U.S. manifest would appear with that name. Unfortunately for us genealogists, they got off the boat and became their old selves again—that is, took back their true names. In other words, without other information, one cannot know whether the name on a manifest was the true name or a purchased name—and, if different from the name used after immigration, the original name or one that simply was adopted!

Sneaking out? One can see how the exit permit requirement led to some illegal emigration. This was made much easier by ticket agents in Germany and others in Marseilles who established smuggling routes out of Russia. Once outside Russia, the emigrants could buy tickets without the permit (since the requirement applied only to tickets purchased in Russia). Given this whole scheme, one can only imagine the information (fact or fiction) the immigrant provided to a ticket agent. These smuggling schemes not only evaded the exit permit requirements, but also other European (mainly German) entry/exit regulations involving medical checks, fees, and so forth. In other words, ticket agents worked as hard to get clients into Germany as out of Russia.

Dependent women and children obviously were not subject to the military service requirement, so the exit permit was not such a hurdle. They needed only ticket money and someone to receive them. They were the perfect candidates for a pre‑paid ticket.

All the above mainly applies to immigration from around 1892 to 1918.

Sack: The old “name was changed at Ellis Island” statement came up repeatedly at my talk. People really did not want to believe that no American immigration inspector ever changed names—although, of course, many immigrants themselves did so after they left Ellis Island. In fact, am I right that a law actually forbade inspectors to do this?

Smith: Not a law, but internal agency guidance. It dictated that no marks or writing were to be made on the manifests except for blue grease pencil marks or writing from the inspector during and regarding inspection (that is, check marks, X’s for detained, etc.); red grease pencil or pen marks/writing made later by verification clerks (annotations for verification, etc); corrections (spelling, name, or other) only as authorized and directed by the commissioner in Washington.

Actually, prior to 1911, no one really cared too much about the names. They wanted them to be accurate, but it was hardly a priority. I cringe whenever I hear anyone talk about how the inspectors at Ellis Island “asked 29 questions.” This is ridiculous, since the inspectors only had a matter of seconds to determine eligibility. Because no legal requirements regarding names existed, any questions on that matter would have been irrelevant. There was no time to discuss names or amend records. Rather, inspectors asked about age, money, jobs, and health.

When the certificate of arrival requirement was enacted in 1911, the Immigration Bureau realized what a nightmare it would be to locate all these mangled names years after arrival. (The Basic Naturalization Act of 1906 included the certificate of arrival requirement, which applied to immigrants admitted after June 29, 1906. These people became eligible for naturalization in 1911—five years later—and so inundated the Immigration Service with verification requests.) From then on, they paid more attention to names, although the system still depended on the steamship companies to prepare the names properly. The law required full and complete manifests. If the names were found to be careless or wrong, that would not be full and complete, and that would leave the company open to fines by the government for each violation. The government could enforce the law more stringently and bring about a change on the part of the carrier. Starting in 1918, the manifests were prepared from passports; from 1917 onward, the manifests were typed, so the names on them should have matched the names (and spellings) on the passports. (Beginning in 1918, the United States required all arriving aliens to present a passport issued to them by their government—or maybe by some other government who for some reason gave them one.)

Missing Ellis Island Cabin-Class Manifests, 1897–1903

In the JewishGen Digest of December 22, 2008, a researcher asked for a “definitive answer” to the question (paraphrased): “Are Ellis Island passenger manifests perfectly complete?” Marian Smith replied on her blog with the following essay:

No, as always, nothing is perfect. But we do have some knowledge to guide us. Records and/or information can be missing for a variety of reasons, some more prevalent in certain eras than in others. The researcher did not give a date for the passenger list he sought, but the biggest question at the moment seems to surround Ellis Island lists at the turn of the century, circa 1897 to 1903, and the presence or absence of cabin-class passenger lists.

The question boils down to the disposition of cabin-class (i.e., first class, second class, and saloon, as opposed to third class/steerage) passenger lists upon the ship’s arrival in New York Harbor. Another writer suggested that cabin-class lists followed cabin-class passengers to the dock where they were discharged and, thus, like the passengers, never went to Ellis Island. Whether the cabin lists ever got to Ellis Island or not, I cannot say. What is clear is that the majority of cabin-class lists for those years were not maintained at the immigration station and were not microfilmed. This is not a myth. It is a demonstrable fact. Before we get to the data, let’s review the collection of passenger lists in New York at the time:

• Prior to 1892, the Customs Service only collected the Federal passenger lists (Customs Lists) and filed them at the Customs House. Presumably, these lists included all passengers in all classes.
• From January 1892 to June 1897 (and somewhat later), the U.S. federal government collected two lists, both of which should have listed all passengers. One went with the customs collector to the Customs House, and the other went with an immigrant inspector to Ellis Island. We do not know if the immigration lists included all passengers or only steerage (third-class) passengers, as no immigration lists from these years survive.
• On June 14/15, 1897, a fire completely destroyed Ellis Island and all the immigration passenger manifests stored there.
• From June 15, 1897, to about January 1903, immigration inspectors continued to collect steerage passenger lists and file them at Ellis Island. They may also have collected cabin passenger lists. If they collected the cabin lists, they did not file them with the steerage records and (as a rule) did not maintain the cabin-class records. Since the cabin lists are not found among Ellis Island records, they may (or may not) have gone to the dock with the passengers.
• After about January 1903, immigrant inspectors collected all classes of passenger lists and filed them at Ellis Island. Although some errors, omissions, and loss continued, for the most part, one can consider the lists since 1903 to be complete in terms of documenting all classes of passengers.
• Following the 1897 Ellis Island fire, the Immigration Service relied upon Customs Lists to verify New York arrivals from January 1892 to June 1897. The New York passenger lists available for those dates at the National Archives and Records Administration (NARA), or online, are in fact Customs Lists.

Why didn’t the Immigration Service collect or maintain cabin lists prior to 1903? We do not know for sure. We do know that from 1892 to 1924 standard Ellis Island procedure included inspection of cabin-class passengers aboard ship and of steerage passengers at the island station. The National Park Service (www.nps.gov/archive/stli/                                                                                      serv02. htm) agrees:

First- and second-class passengers who arrived in New York Harbor were not required to undergo the inspection process at Ellis Island. Instead, these passengers underwent a cursory inspection aboard ship; the theory being that if a person could afford to purchase a first or second class ticket, they were less likely to become a public charge in America due to medical or legal reasons. The Federal government felt that these more affluent passengers would not end up in institutions, hospitals, or become a burden to the state. However, first- and second-class passengers were sent to Ellis Island for further inspection if they were sick or had legal problems.

Part of the government’s rationale was that the majority of cabin passengers were U.S. citizens and returning residents. While I might argue that cabin passengers were not fully inspected until about 1894, that is beside the point. We assume they were manifested, because the government collected and reported statistics by passenger class, and the traditional source for such statistics is the passenger manifest. The New York Times of January 1, 1893 (page 8), printed these statistics from an official Ellis Island statement.

Why do I only assume these statistics came from the manifests? First, because official U.S. immigration statistics have been compiled from manifests (by law) since 1820. Second, I hesitate only because there is a small chance that the numbers could have been provided by the steamship companies. The Transatlantic Passenger Conference (steamship lines) provided passenger statistics by class to the U.S. federal government for the years 1899 to the mid‑1940s. No evidence exists to show that the Transatlantic Passenger Conference collected or reported such data to the government (or anyone else) prior to 1899, but the remote possibility cannot be excluded.¹

By comparing the statistics provided by the shipping lines to actual manifests, one can determine if New York passenger lists of 1897 (or at least 1899) to 1903 are complete. We find that not only are the majority of cabin-class passengers missing from the Ellis Island set in these years, we can (surprisingly) begin to predict the likelihood of finding a given passenger on those lists depending on their port of departure and/or shipping line.

I first compiled the data from the shipping lines regarding New York arrivals for 1899, 1900, 1902, 1903, and 1906.² The 1899 data table below demonstrates the pattern for these years: Except for the Hamburg American’s Union Line sailings from Hamburg, ships sailing from Mediterranean ports had the highest percentage of passengers traveling in steerage. Ships from Continental ports usually had a lower percentage, and ships from British ports had the lowest percentage of steerage passengers. If only steerage manifests survive, it means passengers from the Mediterranean ports are more likely to be found than passengers from British ports, because they are more likely to be listed on a steerage manifest.

I then compared the shipping line data for individual ship arrivals with the digitized records for that ship arrival found on Ancestry.com. The table at the top of the next column shows data for three ships, one each from Britain, the Mediterranean, and the Continent.

The American Line’s St. Louis from Southampton carried 373 passengers, yet only 183 of them appear on the surviving manifest. The one second-class passenger appears only because he is crossed off a steerage list with the annotation “transferred to second cabin.” In contrast, the Fabre Line’s Chateau Yquem from the Mediterranean carried 286 steerage and 3 cabin passengers. All, including the three cabin passengers on their own list, appear on the manifest. Yet even if Ellis Island failed to preserve the page of the cabin-class passengers, the Chateau Yquem’s manifest still would be more than 98 percent complete. Sadly, the Furst Bismarck from Hamburg carried fewer than one-third of her passengers in steerage, leaving more than 532 passenger names out of the surviving records and databases.

The first 1900 example is the Allan State Line’s Sardinian, which arrived at the port of New York on September 27, 1900:

The result in this example is clear. The Sardinian carried approximately 187 individuals to New York, but the surviving passenger list records document only 17 of those passengers—and 17 is the exact number of steerage passengers carried by the Sardinian. Thus, a manifest record survives for only 9 percent of that ship’s passengers, leaving out some 170 names. If Grampa’s naturalization papers say he arrived on that ship, and you haven’t found him among the 17 listed, you can stop looking.

The second ship example from 1900 is the Hamburg American Regular Service. (The Hamburg America Line offered two types of service, regular and express. Express was faster and more expensive.) The shipping lines show that the Georgia arrived on June 12, 1900, although Ancestry.com says that the arrival date was June 11:

The point to note here, aside from the fact that Ancestry found more names than the lines reported (there may have been duplicates), is that all this ship’s passengers traveled steerage and so all are documented.

Further analysis of all the 1900 New York shipping line data shows that ships sailing from different ports offered different accommodations. (See table at <http://web. mac.com/ms5/iWeb/MarianSmith.com/Tables.html>.

Ships from the Mediterranean tended to have all (as above) or the vast majority of passengers booked in steerage. Ships sailing from Continental ports or Britain distributed passengers more evenly among the three classes of travel. As a result, passengers who arrived at New York from a Mediterranean port in these years are more likely to be found on a manifest than those sailing from Britain or Northern Europe, because those sailing from the Mediterranean were more likely to travel in steerage.

If one really looks, reading the lists page by page, one may find a New York cabin-class passenger list from the years 1897 to 1903, but if we assume all the cabin-class passenger lists are missing, the data suggests we are missing nearly 138,000 names in 1900 alone—and passengers who sailed from Britain or the Continent are much more likely to be missing.

The lowest percentage of steerage passengers (0 percent in 1900) was for the Wilson, Leyland, and Atlantic Transport lines—all sailing from British ports. These typically were cattle ships, and the first-class passengers reported principally were cattlemen (U.S. citizens and young British and Irish men). A search of Ancestry.com for 1900 arrivals of Leyland Line ships (Philadelphian, Cestrian, Caledonian, Iberian, Georgian), Wilson Line ships (Buffalo, Colorado, Ontario, Consuelo, Toronto), and Atlantic Transport ships (Marquette, Mesaba, Manitou, Menominee, Minneapolis, Minnehaha) got no hits. In fact, an Ancestry search for any passenger named “John” got no hits on Wilson line ships at New York between April 1897 and August 1903.

Little had changed by 1902. The North German Lloyd ship Friedrich der Grosse from Bremen arrived at New York on August 26, 1902, with 1015 passengers. Only 534 (54 percent) of these were in steerage, more than the manifest pages would suggest. Even if one counts those who did not sail and double‑counts those detained and held, the number of names on the manifest pages still falls far short.

Furthermore, ships from the Mediterranean continued to carry more passengers in steerage than in cabins, in contrast to ships sailing from Britain or the Continent (see tables). The only exception to this general rule was the Hamburg American’s Union Line service from Hamburg, where the ships Pisa, Milano, Barcelona, and Albano carried 100 percent of predominately Russian, Austrian, and Hungarian passengers in steerage. Hamburg American’s Regular Service carried the most immigrants to New York in 1902, but only 84 percent traveled in steerage. As in 1900, those passengers sailing Allan State from Glasgow are most likely to be “missing.”

It would be helpful to have the same shipping line data for years prior to 1899 to see if the same discrepancies noted above appear in ships arriving between 1892 and June 1897 or prior to 1892. Yet, even if we had the data, the inconsistency of 19th-century Customs Lists (missing ships, missing pages) would always leave room for doubt. Nevertheless, I found reports of the New York State Commissioners of Emigration from the early 1870s reprinted by Arno Press, and these contained a few similar tables.⁸ I selected a number of ships from 1871 and their passenger statistics and compared them to the total number of “hits” for those ships in 1871 on Ancestry.com.

The New York State numbers were compiled from Castle Garden (the predecessor to Ellis Island) manifests and are in remarkable agreement with Ancestry’s count of names on the Customs Lists. This lends credence to the conventional wisdom that the Customs Lists are complete, and the omission of cabin-class lists only occurred on Ellis Island, probably beginning in 1892. But because we rely on Customs Lists to mid‑1897, it is only the period June 1897 to early 1903 that lacks cabin-class pages.

Data for the years since 1903 demonstrates that the Immigration Service collected, maintained, and later microfilmed manifests for all passenger classes, despite the fact that a decreasing percentage of passengers traveled in cabin class. For this example, I chose three January 1903 arrivals, one from the three different departure areas.

Beginning in January 1903, the Ancestry.com totals are always higher than or very close to the shipping line totals. This is in marked contrast to the years 1899–1902, when the numbers were always lower unless all the passengers traveled in steerage. This new trend continued in later years, as can be seen in similar data from 1906. For example, the Anchor Line’s ship Astoria arrived at New York from Glasgow, a British port, on November 23, 1906. The adoption of government‑issued forms for different classes of passengers allows us to count passengers on the manifest pages according to class.

In the 1906 Astoria case, the data matches almost exactly. The difference of two is probably explained by Ancestry.com’s indexing of passengers who did not sail. A second 1906 example reinforces the proposition that post‑1903 New York lists are complete.

Missing Names

The results of this simple study suggest that the vast majority of passengers who arrived at New York from January 1903 onward on a regular commercial carrier should be documented on a passenger list (even if illegibly). Those who cannot be found since 1903 may be persons traveling under an assumed name, including children traveling with relatives under another family name. Or they may have been seamen who jumped ship prior to May 1, 1917, and so are not documented in crew lists.

The data also shows that Ellis Island records from at least 1899 (and presumably June 1897) to early 1903 are comprised of steerage passengers only, with few exceptions. Hence, as a rule, passengers who traveled in first or second (cabin) class will not be found. Furthermore, the passenger data shows that ships sailing from the Mediterranean tended to carry most or all of their passengers in steerage and thus increased the possibility of that passenger’s name surviving on a record today. Most ships from British and Northern European ports carried a lower percentage of passengers in steerage, decreasing the chance of finding a particular passenger on a surviving list. Passenger data of 1903 and 1906 agrees almost exactly with corresponding manifest records, demonstrating that all classes of passenger lists can be found on the microfilm after January 1903—(even if every immigrant cannot!).

Sack: So to go back to the original question—Why do you think so many immigrants said that “the name was changed at Ellis Island?”

Smith: My best guess is that the words “Ellis Island” were shorthand for the entire immigration process. Somewhere along the way from the old life in Europe to the new life in America, the immigrant acquired or adopted a new name. Just exactly where and how probably was not the same for every immigrant, but one thing we do know—the name was not changed at Ellis Island. The other interesting phrase is “government official,” as in your family story that some “government official” changed your relative’s name in South Africa. To an immigrant, anyone in a uniform might have seemed like a government official. Thus, if the ticket agent from whom a steamship ticket was purchased wore a uniform, then he likely would have been considered a “government official.” We need to try to put ourselves into the mindset and lives of our immigrant ancestors in order to understand their experience.

Notes

1. See Drew Keeling. “Repeat Migration between Europe and the United States, 1870‑1914,” Institute of European Studies, University of California Berkeley, Paper 050411 (2006), App. 1.
2. Transatlantic Passenger Conference data, 1899, 1900, 1902, 1903, and 1906. USCIS Historical Reference Library, 111 Massachusetts Avenue, NW, Room 1000, Washington, DC 20529.
3. Ancestry.com has two pages for this ship arrival, filed under the dates September 21 and 27.
4. When counting passengers from manifest pages, I ignored names that are crossed‑out or noted as “did not sail,” as well as names on detained and Board of Special Inquiry lists, since these entries would have been omitted from the shipping line count.
5. Count from manifest pages, minus passengers who did not sail.
6. Count from manifest pages, including lists of aliens detained or held for special inquiry (duplicates).
7. Count from the Ancestry.com database of hits for this ship.
8. Friedrich Kapp, Immigration and the Commissioners of Emigration, (New York, NY: Arno Press and the New York Times, 1969).