In 1921 the remaining Royal Hanoverian Creams at Hampton Court were auctioned off. These horses had pulled the English State Coach since the reign of George I. The majority of the stock ended up in circuses; but despite at least one breeding program there was no success in continuing the breed. Unfortunately, we have no color photographs and no living descendants of the Royal Hanoverian Creams, and so there has been some debate as to what color the “Creams” actually were. What we do have are a large number of paintings and descriptions, a handful of black and white photographs, and one wooden statue. From these primary sources, I have compiled a set of identifying traits. These characteristics are: diluted body color that, while paler than normal coat colors, retains pigment; mane and tail that is often darker than the body, yet still dilute; commonly, leg shading that is fairly consistent from hoof to above the knee and hock; diluted skin tone, particularly as shown around the muzzle; and eyes that were either brown, amber, or possibly even green—indicating a dilution of the eye color. The way newspaper articles of the 19th century describe these horses suggests that these traits were novel, several articles even suggesting that the exotic color stemmed from ‘oriental’ imports.
The obvious assumption is that the Hanoverian Creams were, in fact, cream. There is a partial dominant dilution gene in horses that is called cream, and double-dilutes (horses with two copies of this gene) are categorically called cream. There are several lines of warmblood horses that carry the cream gene and claim the source to be the Royal Hanoverian Creams. None of these have records of this fact, but rather use it as justification when a dilute foal appears suddenly, as can happen with the cream gene due to the fact that it does not effect black pigment when heterozygous. However, there are a handful of problems with this theory. First, cream dilutes have a self-colored mane and tail more often than darker, and leg shading is rare. Secondly, the cream dilution is a partial-dominant trait: when a double cream dilute is bred to a non-dilute horse, the resulting foal will have neither parents color; instead, they have an intermediate shade. The Hanoverian Creams, by contrast, consistently failed to produce and form of dilute colored foal when outcrossed. And finally, so called “blue eyed creams,” being double cream dilutes, were deemed unhealthy albinos by every European horse registry during the same period that the Hanoverian Creams were in use.
The second option is a dilution known as champagne. Champagne fits the phenotype (the set of characteristics I listed earlier) reasonably well, with the small exception that champagne tends to cause heavy mottling of the skin. This trait, however, is very unevenly expressed and might not be reproduced in painting or easily seen in old black and white photos. A breeder in Tennessee has attempted to re-create the Hanoverian Cream, primarily using the American Cream Draft. The American Cream Draft, like the Hanoverian Cream, has a misleading name. They are, in fact, champagne. It has been suggested that the progenitor of the American Cream Draft, a mare named Old Granny bought at an auction in Iowa in 1913, was one of the dispersed Hanoverian Creams. You may have noticed that the dates don’t quite match up, as the dispersal sale was not until 1921. Although letters from the Master of the Horse at the time suggest that the herd at Hampton Court was slowly thinned in the years leading up to the auction, there is no mention of any of breedable Creams being sold to the public before 1921. There were at least two Hanoverian Creams in the possession of Phillip Astley’s Royal Circus in 1850. And, King George III did sell a small number to the Hanoverian Creams to a London cab company almost a century earlier, possibly in fury at Napoleon’s seizure of the stud farm in Hanover. Despite this possible source of Hanoverian Cream bloodlines outside the royal studs, it remains very unlikely that the Creams were champagne. Both the circus horses and the cab horses were most likely geldings. And, since the recent advent of a genetic test, no champagne horses have been found in Europe. Based on the distribution of the gene, it likely originated in the United States in a light horse breed in the early to mid 1800’s. The Hanoverian Creams, by contrast, are documented to about 1700.
That brings me to the final option: pearl. Many of the horses in Europe thought to have been champagne have since tested positive for pearl, a mutation that was not even theorized to exist until the 90’s. Pearl closely fits the phenotype of the Hanoverian Creams. It is also a simple recessive trait, which explains the lack of success in outcrossing. And, pearl has been identified as stemming from Iberian horses, which is unsurprisingly the likely origin of the Hanoverian Creams. Despite the orientalist rhetoric attached to the Cream’s unique color in the 19th century, the myth surrounding their pedigree suggests the original stock made its way to Hanover via the annexation of Prussia, taking over stock said to have been a gift from Isabella of Castile. A common term for a variety of cream colors in horses, birds, and even fabric in the 17th to 19th centuries was ‘isabel.’ The OED defines ‘Isabella’ as: “Greyish yellow [or] light buff” and states that “various stories have been put forth to account for the name. That given in D'Israeli Cur. Lit. … and also in Littré, associating it with the Archduchess Isabella and the siege of Ostend 1601–4, is shown by quot. 1600 to be chronologically impossible.” Interestingly, there is an oral tradition among equestrians of several countries that attributes the term (used for a variety of dilute coat colors) not to the Archduchess of Austria but to Isabella of Castile, who is purported to have had stable full of ‘golden’ horses. She would fit remarkably well with the distribution of the word ‘isabella’ for a golden coat, which is used in English, French, German, Dutch, Russian, and others throughout the early modern period and to some extent still today; and also fits the distribution of both pearl and cream horses.
Most interesting to me, however, is the descriptions and paintings of the Hanoverian Cream’s heads and necks. They are commonly shown with a sloped shoulder; a high, thick, arched neck; occasionally a convex skull, and never a concave skull. These traits are so unique to the Iberian breeds that they were grounds for multiple genetic studies to determine if they had been a unique species with a separate domestication point. Whether or not Isabella of Castile was involved, it is almost certain that the Hanoverian Creams had a considerable amount of Iberian influence, which only makes it more likely that they were double pearl dilutes.
To complicate matters, the genes responsible for cream and pearl occur at the same locus. This means that a horse can be homozygous for either cream or pearl, negative for both, or heterozygous for both, but never homozygous for one and still carry a copy (or two) of the alternate dilution. Pearl, normally a recessive gene, is also “cream activated.” A horse that is heterozygous for both cream and pearl will have a double dilute phenotype that generally appears slightly darker than a double cream, but slightly lighter than a double pearl. They will be easily distinguished from double cream dilutes by the production of non-dilute (though pearl carrying) offspring from non-dilute mates. However, when cream-pearl is bred to cream-pearl they will appear to breed true, consistently producing dilute foals. The resulting foals may be double cream, double pearl, or cream-pearl, but all will exhibit ‘goldish’ coats. Because of this, along with the etymological distribution of ‘isabella’ and relative distribution of both colors, it is likely that some of the Hanoverian Creams were cream-pearl, and therefore some may also have been double cream dilutes. An 1833 German breed book notes that there are occasional abnormalities “ unter den Isabellen kommen sehr lichtgefärbte Pferde vor, bei denen der dunkle Streif auf dem Rücken fehlt, ... die Iris ganz weiss ist, ohne schwärzlichen Rand, und auch die Traubenkörper eine weissliche Farbe haben,” which describes the double cream dilutes. This would explain the occasional blue-eyed and pale-pointed painting. However, considering the overwhelming number of paintings that do not exhibit these traits, the photographs of ‘Pistachio’ (who, being one of the last sold, was likely ideal), and the lack of success in producing like colored foals from outcrosses, the cream-pearls were likely a rare exception.
 The category is cream, but there are individual terms depending on base coat color. These are: Cremello (2 cream on red), Perlino (2 cream on bay), Smokey Cream (2 cream on black), and as yet 2 cream on brown does not have a consistent name.
 warmblood is a middle weight breed category, and has not relation to actual blood temperature.
 An unhealthy animal was likely to be sent as “a gift to the kennel” at a young age. Sporting, V6, 225.
 Pearl arguably does express when heterozygous, but so mildly that even the educated eye has difficulty identifying it definitively without the aid of pedigree.
 “among the Isabel occur very light coated horses, where the dark stripe missing on the back ... the iris is completely white, without blackish edge, and the eyelashes have a whitish color.” It may be argued that the normal presence of a dorsal stripe would indicate dun. However, none of the Hanoverian Creams had the facial masking typical of dun, and pearl will often have a less diluting effect on a countershaded dorsal. Seiler, 56.
 ‘points’ in horse color are the legs, mane, and tail. Recall that pearl tends to leave the points darker, especially on non-red based coats.