82 INHERITANCE IN POULTRY. 



obscurely barred with dusty bands, althougli, on the other hand, this barring 

 may be merely the badge of hetero/.ygotism. It is possible, therefore, that 

 the barring in the plumage of the White Leghorn Bantams is transmitted as 

 an alternative characteristic. The case of the barred descendants of the 

 Tosa X White Cochin is more difficult. I am not yet prepared to go so far 

 as Correns (1905"*, p. 13, note) when he says: "Wo Mosaikbildung als 

 Regel bei einem Bastard auftritt, war sie scbon in einem der Eltern oder in 

 beiden, aktiv oder latent, vorhanden," 



Naturally, attention was directed chiefly toward evident qualitatively 

 marked characteristics. Such do not blend. The fact that for the most 

 part a characteristic does not blend when crossed with its allelomorph is of 

 the highest importance for the theory of evolution. If blending were uni- 

 versal a new characteristic must inevitably become quickly swamped by 

 intercrossing with the parental characteristic. Since the new quality does 

 not blend, it need not be swamped, even when there is no special isolation. 



INHERITANCE OF SPECIFIC VS. VARIETAL CHAR.\CTERISTICS. 



A distinction between specific and varietal characteristics is made by 

 Nageli (1884, p. 247) and by de Vries (1902, p. 141 ; 1905, p. 141). Fol- 

 lowing de Vries, a specific characteristic is a wholly novel one acquired by the 

 race — one which stamps its possessor as an elementary species, A varietal 

 characteristic is sometimes positive (/. <?., additional), in which case it is 

 found also in closely allied species, and may be regarded as the becoming 

 patent of a characteristic all the time latent in the variety. It is, on the 

 other hand, sometimes negative, this condition being marked by the disap- 

 pearance (becoming latent) of a characteristic patent in the ancestral species. 

 Specific and varietal characteristics are thought by de Vries to be inherited 

 very differently. When two elementary species are crossed the character- 

 istics of both parents appear, fully developed, side by side ; Mendel's law is 

 not follov/ed. When a species is cro-ssed with a variety a Mendelian result 

 is obtained and the patent characteristic is dominant over the latent. 



Let us now see in how far the results gained in breeding poultry accord 

 with de Vries's law. It is not easy to make the classification in an unpreju- 

 diced way ; an attempt, however, will be made. 



First, the comb is a specific characteristic of the genus Gallus. It is absent 

 in other Gallinse. Also pea comb and rose comb are each wholly new, posi- 

 tive variations from the primitive single comb. Muff and beard seem to be 

 novel ; so also the long tail of the Tosa fowl, the extra toe, and the melanic 

 feet and beak. 



Clear cases of negative variations are : Loss of the nasal process of inter- 

 maxillary and consequent high nostril ; failure of cerebral plate to close and 

 consequent cerebral hernia ; loss of uropygium ; loss of red and black pig- 

 ment in feathers (albinism, partial or complete) ; loss of dark pigment in 

 crest feathers ; loss of wing bar ; loss of primitive shafting. 



