156 
necessarily followed by reversion. The offspring, or some of them, may (1) down 
to the smallest details resemble one of the parents, sometimes the male, sometimes 
the female; or (2) they may reproduce both the mental and physical peculiarities of 
one of the grand parents, or of even a fairly remote ancestor; or (3) they may con- 
sist of an unequal mixture of two or more Breeds and well deserve to be designated 
mongrels; or (4) they may combine the more striking characters of the two breeds; 
they may present quite new characters; when this happens, they are often 
termed 'sports.' 
Ewart further calls attention to the fact that in large litters of dogs, cats, pigs, rab- 
bits, etc., sometimes both parents and one or more remote ancestors are faithfully 
reproduced, and when the parents belong to fairly distinct strains, there may be 
almost a restoration of a quite reunite ancestor. He gives an illustration from a pho- 
tograph of three rabbits, full sisters. One reproduces the characters of the dam, one 
an Angora like the paternal grandam, and one a Himalaya like the maternal great- 
grandam. Of a litter of four kittens from white parents, two are white like the 
parents, and two dark colored like the great-grandam. 
In a discussion of telegony, Ewart picks to pieces the evidence concerning the 
Lord Morton quagga hybrid and throws great doubt upon the original interpretation. 
Then coming to his own investigations, he says: "1 may mention that neither in the 
case of horses, cattle, sheep, dogs, cats, rabbits, mice, or guinea pigs, nor yet in the 
case of ducks, fowls, or pigeons, have I ever seen a ease of telegony." Many remark- 
able instances of variation and reversion (which some would doubtless have explained 
by the infection hypothesis) have been observed, but in every instance it was possi- 
ble to account for the phenomena without resorting to the telegony doctrine. He 
then refers to his experiments with horses and cites ten mares that have had an 
opportunity of being infected by Matopo, the zebra. There is not the slightest 
evidence that any of these mares have ever been infected. 
The methods of Ewart are of the new school, so far as investigating breeding prob- 
lems are concerned, and should inspire others to take up similar research in this great 
field. 
Before passing to another phase of this subject it will be appropriate here to refer 
to some of the experimental work on breeding conducted by an American institu- 
tion. I refer to the poultry-breeding experiments of the Rhode Island Station, and 
more particularly those relative to geese and to some extent turkeys. For years the 
poultry industry has received attention by the Rhode Island College and Station. 
Inasmuch as Rhode Island has considerable poultry interests and is famous for her 
turkeys, it was quite appropriate for the college to promote this industry. The vari- 
ous reports published by the station contain in the aggregate much interesting and 
valuable information. 
In 1892 the station undertook its first special breeding work by introducing a 
wild turkey gobbler, raised from wild turkey eggs from Maryland, and attempted 
breeding him to domestic hen turkeys. The purpose in this was to introduce enough 
wild blood in the domestic turkey to secure some of the hardiness and vigor of the 
wild stock, and thus add to the vitality of domestic birds. The. first season this work 
was attempted the wild gobbler would not associate with the domestic hens, drove 
them from him, and did not strut or gobble that season. The following year 4 
wild gobblers were secured, and these were used by the station, and several farmers 
also engaged in the work. From this stock a quantity of cross-bred turkeys were 
produced by the station, which were kept under observation. A number of Rhode 
Island farmers also produced some crossbreds. Following this, the next season, the 
station distributed three-eighths and half-bred wild gobblers to 42 Rhode Island 
turkey growers for practical breeding and observation in their flocks. It was con- 
sidered that birds resulting from this cross improved the turkey stock of those 
using it. One person who reared 300 turkeys from three-quarter wild gobblers secured 
a larger percentage of mature birds for market, and they proved more uniform in 
size, larger in weight, fatted quicker, and dressed better than those of domestic blood. 
The most elaborate breeding work by this station, however, has been with geese. 
This began in 1893. Toulouse and Einbden geese were crossmated each way, and 
records kept of the development of the goslings from June 19 to September 5, weekly 
weights being made of the crosses, as well as extended records of pure Toulouse and 
Embden stock. In L894 cross-breeding work on a somewhat larger scale was under- 
taken, and African and Chinese crosses were also introduced. Records of the growth of 
the products of the various crosses, 52 birds in all, were made, showing the growth at 
various dates from June 9 to December 8, 1894. Later specimens were slaughtered 
and a record made of live and dressed weight, with shrinkage, and the dressed birds 
a Guide to the Zebra Hybrids, p. 25. 
