By 1868, Nous had been an established working retriever at Guisachan for three years, and his owner, the 1st Baron Tweedmouth (Dudley Marjoribanks), decided that he wanted to use Nous to found his own breed.
Dudley Marjoribanks had grown up in Berwickshire, the former county in which Berwick-upon-Tweed had been its shire town. He also represented Berwick-upon-Tweed as an MP, and thus, he was familiar with that region’s peculiarities.
He knew of the local water dog, which was a cross between the indigenous water spaniel of the region and “the Newfoundland.”
Richard Lawrence wrote about them in The Complete Farrier and British Sportsman in 1816:
Along rocky shores and dreadful declivities beyond the junction of the Tweed with the sea of Berwick, water dogs have derived an addition of strength, from the introduction of a cross with the Newfoundland dog, which has rendered them completely adequate to the arduous difficulties and diurnal perils in which they are systematically engaged (405).
These dogs were a landrace type, which means they varied greatly in appearance. In Hugh Dalziel’s British Dogs, J.S. Skidmore’s description of the Tweed water spaniel goes as follows:
They were very light liver colour, so close in curl as to give me the idea that they had originally been a cross from a smooth-haired dog; they were long in tail, ears heavy in flesh and hard like a hound’s, but only slightly feathered – fore legs feathered behind, hind legs smooth, head conical, lips more pendulous than M’Carthy’s strain. The one I owned, which was considered to be one of the best of them, I bred from twice, and in each litter several of the puppies were liver and tan, being tanned from the knees downward and under the tail. I came to the conclusion that she, at any rate, had been crossed with the bloodhound.
It is possible that his dog had been crossed with bloodhound or maybe Gordon setter, but at least one account of the dogs suggests that at least some of these dogs were more of the retriever-type Stanley O’Neill was a well-known flat-coat expert who he had encountered Tweed water dogs as a boy in the 1890’s. His description is of a more retriever like than that of J.S. Skidmore:
Further up the coast, probably Alnmouth [in Northumberland, south of Berwick-upon-Tweed], I saw men netting for salmon. With them was a dog with a wavy or curly coat. It was a tawny colour but, wet and spumy, it was difficult to see the exact colour, or how much was due to bleach and salt. Whilst my elders discussed the fishing I asked these Northumberland salmon net men whether their dog was a [St. John's?] Water-Dog or a Curly, airing my knowledge. They told me he was a Tweed Water Spaniel. This was a new one on me. I had a nasty suspicion my leg was being pulled. This dog looked like a brown Water Dog to me, certainly retrieverish, and not at all spanielly. I asked if he came from a trawler, and was told it came from Berwick.
From that description, the dogs looked like a tawny curly-coated retriever. This suggests that at least some of the dogs were not true livers but were brown-skinned yellow to reds. The “light-liver” color in the Skidmore description sounds more like a deadgrass Chessie than a true liver-colored dog. (Deadgrass Chessies are light yellow dogs with brown skin.)
Now, from my reading of all of these texts, a Tweed water dog or Tweed water spaniel was actually a derivative of the St. John’s water dog. That is why it looked so much like a retriever. The fact that the dogs had such short hair suggests that they were derived from that “Newfoundland,” rather than the big one. It is likely that the native water spaniel in Northumberland and the Borders was red or yellow in color, rather than truly liver.
Also, in the O’Neill description, the dogs were being used to net salmon. That particular job is the exact task that the St. John’s water dogs performed in Newfoundland.
The dogs were celebrated waterfowl dogs, retrieving shot birds from the chilly and rough waters of the North Sea coast. Because this was a regional breed, it was not well-known in rest of Britain. Dudley Marjoribanks most likely knew about them and their reputation as superior retrievers.
However, in those days, the preferred color for a retriever was black. Other colors simply were not bred from. Perhaps Marjoribanks’s experience with Nous and his knowledge of the Tweed water dog gave him enough confidence to challenge the accepted wisdom of the day.
We do not know what Belle, the Tweed water dog chosen as Nous’s mate, looked like. We can only infer from the depictions of their offspring.
Nous appears to be rather dark-colored dog that was somewhat heavy in bone. If you saw him today, you would recognize him as a golden retriever.
Ada, Crocus, Primrose, and Cowslip, the four bitch puppies that resulted from that breeding, also looked a lot like goldens. Two depictions of those puppies exist– one of Ada and one of either Cowslip or Primrose. Ada is a rather short-haired dog. The dog said to be Cowslip or Primrose has rather wavy long hair.
Both of these dogs are lighter in color than their sire, and both are more lightly build than their sire. This suggests that Belle was a more lightly built dog than Nous and was of a pale gold color. The paler shades in the golden retriever most likely come from the Tweed water dog, for the red t0 yellow wavy-coated retrievers and red Irish and Gordon setters that were crossed into the strain are not that pale in color.
Belle was most likely a brown-skinned yellow, while Nous was a black-skinned yellow of the darker shade.
So now we have an idea about what the two foundational breeds that helped create the golden retriever looked like. You can see some of the Tweed water dog’s characteristics in some golden retrievers, especially in the performance-bred lines. This breed disappeared by the turn of the century, mostly by being absorbed into the retrievers. Regional dogs also had a hard time competing against the “improved” breeds of retriever that were coming to the fore as the nineteenth century progressed.
There are two often postulated theories about the origin of the St. John’s water dog, and I just don’t think it’s probable.
1. Richard Wolters believed they were derived from the black St. Hubert hound. He claims that the St. Hubert hounds were used to retrieve from a from Turberville that says the St. Hubert hounds were commonly “blacke” and used for finding “farre straggled” game. I’m surprised Wolters latched onto that theory, because he apparently didn’t know that St. Hubert’s hound is what was called a “lymer.” Lymers were used to track wounded game, which is actually one theory about how the bloodhound got its name. It actually trailed blood spoor. Wolters thought that this was retrieving. It’s not.
2. Another theory goes that these dogs are derived from the French matin dog. A matin is an extinct dog that had features of the greyhound and mastiff– something like a Great Dane. It also was probably derived from the alaunt veantre. It is not and never has been a water dog. As far as I know it didn’t retrieve. Not a very good candidate. When I first read about this theory, I had just read a description of the French matin. I don’t remember who came up with this theory, but the water dogs are no more likely to have been derived from this dog than they are from pekes or Afghan hounds.
The best and most logical theory is that these dogs are very closely related to the poodle type. Yes, even the big Newfoundland dogs most likely originally derived from something like a poodle. The most likely candidates are the Portuguese water dogs, Spanish water dogs, and the English rough water dog. These may have been interbred with the collie-types, setter and pointer types, water spaniels, and the Cao de Castro Laboreiro or something very similar to it. It is also possible that the large Arctic spitz from Labrador called the Labrador husky played a role. And it is very likely that some form of mastiff, probably the English mastiff, played a role in developing the larger strain of Newfoundland that was used mostly for hauling loads.
I am skeptical that the Great Pyrenees or its closer relatives played much of role, simply because the dogs lived deep in the mountains, far away from the coastal Basques who were visiting Newfoundland. And while a smaller guard dog, like a Cao de Castro Laboreiro, might be of some utility, a big one would be too costly in terms of the amount of food it would need to survive. Bringing a livestock guardian dog of that size across the Atlantic just to use as a watch dog simply wouldn’t have been economical, and these dogs were more valuable in the Pyrenees as flock guardians.
The best theory as to why these dogs lost their poodle type coat can be found here. And evidence of poodle-type water dogs in Newfoundland in the early days can be found here.
Farley Mowat traced his St. John’s water dog’s ancestry to the poodle-type dogs. In fact, he had reason to believe that they were related to the Portuguese water dog, which is often black with white markings– often exactly the same sort of white markings than were found on the St. John’s water dogs. It also occasionally has puppies that have coats a bit like a flat-coated retriever, which then would be very easily developed into a smooth-coated water dog. The Portuguese water dog’s temperament is very similar to that of a retriever. They are very friendly and eager to please. And many have very strong retrieving instincts.
Now, Mowat also traced the poodle-type dog’s origins all the way back to Russia. Now, here I disagree but only slightly. I think Central Asia is where they actually came from. The puli is a relative of this type and is a good example of what these dogs originally were. They are were herding dogs. I can forgive Mowat, though, for at this time, most of Central Asia was under the control of Soviet Union.
I don’t know where people get these theories. I think it may be disconcerting for poodle-haters to accept that these all of their hunting dogs related to dogs of this type. They have be derived from hounds or mastiffs. In the case of these St. John’s water dogs and the Newfoundland, it is very common for people to hold onto the tired old story that Leif Ericsson brought them over. How these dogs managed to survive on their own for the nearly 500 years before the modern period of European colonization is a good question. They would have had to have survived on their own, because the Beothuks didn’t keep dogs.
Whatever, the reason, the fact that Portuguese water dogs are marked so similarly to the St. John’s water dog and posses the playfulness, retrieving instincts, and strong swimming ability associated with these ancestors of the retrievers means that we should look at them more closely. The Portuguese used the Grand Banks extensively until the end of the sixteenth century. The Portuguese, Spanish and French basques, and the English used Newfoundland as a base for fishing, and it is very likely that the water dogs from Portugal became common on the island, even after their original owners gave up on the fishery. Thus, the Portuguese water dog really should be carefully considered as a forebear of the St. John’s water dog, the Newfoundland, and the retrievers.
That Portuguese water dogs don’t often appear in duck blinds is immaterial.
Their conformation, temperament, and working instincts very strongly suggest that they are most likely the main ancestor of the St. John’s water dog, although I won’t rule out the Spanish water dog (which is actually used more often as a herding dog), the English rough water dog, or the barbet.
But it is clear to me that the St. John’s water dogs, the big Newfoundland, and the retrievers belong in this family. That they lack the non-shedding coats and the beards is immaterial. There aren’t that many genes that control the type of coat that a dog has, and it wouldn’t take very long for a poodle-type water dog to develop a smooth coat through selective breeding, natural selection, or both.
