The Tide That Turned Both Ways

The Challenge

David Paterson Robertson, a mason’s son born September 1842 in Blairgowrie, Perthshire, married Elizabeth Gray of Monymusk, Aberdeenshire, in Dundee on June 23, 1866. Their first son, William, was born in Dundee in April 1867. By the 1870 census the young family was settled in Brooklyn’s 12th Ward — David working as a stone cutter near the waterfront, two American-born daughters in the household. The 1874 Brooklyn city directory places him at 119 Hamilton Avenue, the address of his recently widowed mother. A son, George Paterson Robertson, was born at that same address in May 1874, his Brooklyn birth return naming both parents and their Scottish birthplaces.

Then the trail stops. After 1874, the family disappears from Brooklyn’s directories and vital records. When the directory listings resume in 1880, David Robertson, stonecutter, is back — at 119 Hamilton Avenue, exactly where he left off.

A five-year silence in American records is usually read one of three ways: the family moved west, the family died, or the enumerator missed them. None fits. They did not appear elsewhere in the United States. They demonstrably did not die. And a directory gap of five consecutive years is not an enumeration accident — directories were compiled annually, by canvass.

The instinctive next step — find the passenger lists — fails by design. The United Kingdom did not keep inward passenger lists for arrivals from outside Europe and the Mediterranean until 1878. A family sailing from New York to the Clyde in 1875 generated no surviving arrival record at all. The peak years of return migration from depression-era America, 1873 to 1878, are precisely the years for which no direct evidence of arrival can exist.

The question, then, is the genealogist’s hardest kind: can a transatlantic migration be proven when the migration itself left no record? The Genealogical Proof Standard says yes — if the indirect evidence is independent, convergent, and correlated. This case is the demonstration.

The Breakthrough

This is the feature that makes the case provable at all. An American birth certificate of the period names parents; a Scottish statutory entry anchors them to a specific marriage event. The mason’s occupation carries through every record — stone mason in Dundee 1867, stone cutter in Brooklyn 1870 and 1874, mason (journeyman) in Maryhill and Govan — the same trade, rendered in each country’s idiom.

The 1874 Brooklyn directory lists David Robertson, stonecutter, at 119 Hamilton Avenue; George’s May 1874 birth return places the family there in residence. The 1880 directory lists David Robertson, stonecutter, at 119 Hamilton Avenue. Between those brackets sit the two Glasgow births. The family did not scatter, fail, or vanish — it crossed an ocean and came back to the precise point of departure.

Decades later, the U.S. census quietly corroborates the whole structure. In 1900, David reports immigration in 1870 — his first arrival. His son William, enumerated separately as a Brooklyn mason, reports birth in Scotland and immigration in 1878 — the re-emigration year. William’s 1910 census entry repeats it: born Scotland, immigrated 1878. A child of two Brooklyn-settled parents, born in Glasgow mid-decade and arriving in America in 1878, is inexplicable without the return.

Context explains what the records prove. The Panic of September 1873 collapsed American construction — the stone cutter’s entire market — while Scotland’s iron and coal boom drove Glasgow wages to unusual heights. Tahitia McCabe’s 2017 study of the 1881 Scottish census found 2,167 American-born residents of Scotland, more than half of them children under fifteen, and traced over three-quarters of their parents as Scottish return migrants — concentrated in the Glasgow industrial belt, with construction among the leading occupations. The Robertsons match that documented wave point for point. The pattern does not prove their journey; the records above do. The pattern explains why the journey made sense.

The Result

David Paterson Robertson and Elizabeth Gray emigrated from Scotland to Brooklyn around 1868–69, returned to the Glasgow area about 1875, and re-emigrated to Brooklyn in late 1878 — a circular migration established entirely through correlated indirect evidence, because no direct record of the eastbound crossing can exist. Two statutory birth entries anchored to the couple’s 1866 Dundee marriage place them in Maryhill and Govan; matching directory listings at 119 Hamilton Avenue bracket the absence; and the census immigration years — 1870 for the father, 1878 for the Scottish-born son — encode the round trip in the family’s own later testimony.

Elizabeth’s Scottish origins stand alongside it. Her 1846 Monymusk baptism, her 1861 census entry at Newcroft as a crofter’s daughter, the three statutory records naming her as Gray, and the death records of her children consistently stating “Scotland” together outweigh — and explain — the single erroneous “U.S.” notation of 1884.

The 1900 census records the family’s stark arithmetic: eleven children born, six living. Nine are now identified — including both children of the Scottish sojourn. Neither survived to adulthood except William, the Maryhill-born son who carried his Scottish birthplace through every later American record and outlived all his siblings, dying in 1948. Joseph’s 1884 certificate, for all its misrecorded surname and birthplaces, supplies a corroborating detail: it numbers him the mother’s tenth child.

The family the return preserved went on to its own dramatic arc. Elizabeth died in Brooklyn on July 4, 1902 — born on Christmas Day, dead on Independence Day. David, widowed, left stone cutting around 1906 for the swamps near Savannah, Georgia, where he worked as a game trapper until February 1910, when his boat was found swamped and he was never seen again. His son Joseph’s newspaper appeal — “Is David Robertson In or Around Macon?” — ran that October. The full story is told in Episode 6 of the Scattered Stones series.

Return migration is one of the most underdiagnosed phenomena in Scottish-American genealogy, precisely because its peak years fall in the pre-1878 documentation void. If your Scottish family disappears from American records in the mid-1870s, they may not have moved west or died — they may have gone home. Look for American-born children in the 1881 and 1891 Scottish censuses; for Scottish statutory births citing a marriage that took place before the family emigrated; for directory listings that stop and restart at the same address; and for census immigration years that differ between parents and children within one household. Each is indirect. Correlated, they prove what no passenger list survives to show.