218 THE ZOOLOGIST. 



become collateral to a far greater extent than in a small stand, 

 and the effect of individual variability would make itself more 

 felt. 



If a large stand is examined, which we will presume has been 

 descended from a first-class coloured Italian strain, and to have 

 been kept unmixed for several years and allowed to increase by 

 natural swarming, it will be found that nearly all colonies vary 

 more or less in colouring, and probably more still in other 

 characteristics. No doubt some extremely fine-coloured colonies 

 will be met with, but the greater number, as a rule, do not come 

 up to the desired excellence. Some cases may perhaps be 

 noticed that are not much better than a cross between a Brown 

 and an Italian bee, and, on the whole, a tendency towards a dark 

 colour is perceptible. 



It may be advanced that, as the Ligurian bee is a secondary 

 race produced by a crossing of the Brown and the Egyptian race, 

 the Ligurians are in reality hybrids, and that consequently it is 

 not a very weighty argument to apply the test of reversion upon 

 these bees, as hybrid blood is more subject to reversion than that 

 of a purely-bred variety. Admitting this fact, it must, however, 

 be remembered that the deterioration of the Ligurian bee, which 

 is the same as reversion in this instance, is always stronger 

 towards the dark colour than towards the light, whereas in a true 

 half-blood there should be no greater tendency in the one 

 direction than in the other.* 



Moreover, as has been shown, the disposition to vary in the 

 direction of coloured bands prevails among all varieties of the 

 Brown bee, and must, therefore, be considered an innate charac- 

 teristic of its organisation, and consequently, according to the 

 principle of the frequent accumulation of a variation, should 

 rather assist than check the perpetuity of the yellow bands. 

 That such is at times the cases is shown by the crosses between 

 the Brown and the Ligurian bees, which when allowed to breed 



* The reversion to an ancestral type must not be confounded with the 

 reversion of a half-blood to a characteristic possessed by one or other strains 

 of the cross. Such a reversion is of common occurrence, but it has not been 

 observed to happen after many generations, and certainly ceases to appear 

 beyond the twentieth generation. This kind of reversion, therefore, cannot 

 apply to the Ligurian bee, which was produced at least two thousand years 

 ago. 



