MASTERS' BOOK ON VEGETABLE TERATOLOGY. 



161 



construction. Other examples may be cited, and the ironwood, or Casua- 

 rina (fig. 46, a), which is in so many respects like the horse-tails, is 

 sometimes seen, in our conservatories, to twist in the same way. Partial 

 spiral torsions, of course, also occur in the teasel and often give rise to 

 additional anomalies. 



Spiral torsion may be considered as a case of complicated atavism, 

 caused mainly by the loss of the position of the leaves in pairs or in 

 whorls. They return to the more primitive spiral arrangement. 



This is the main point, and all the other striking deviations may be 

 considered as the mechanical consequences of this primary change. For 

 the evolutionist, this change from whorls into the spiral condition gives 

 proof of the development of the more specialized arrangement, from the 

 undifferentiated, and more common, spiral arrangement. 



Laciniate leaves, in some instances, afford examples of correlation. 

 This is a most interesting question, and one to which Masters has given 

 much attention. It gives us an insight into the internal causes of 

 monstrosities. The same anomaly may betray itself in different organs, 

 provided they are homologous. Such is evidently the case with petals 

 and leaves, and from this we may expect to find correlations in their 

 anomalies. Laciniate leaves induce laciniate petals, and the corre- 

 sponding variety of the ordinary bramble is one of the best known 

 instances (fig. 47). 



Pitchers (fig. 48) are, ordinarily, rare abnormalities. In the Magnolia 

 they repeat themselves regularly, but in other plants they are seen but once, 

 or at long intervals only. Some of them are formed from one leaf, others 

 from the two leaves of a pair ; and still others from the leaflets of 

 compound leaves, as in the clover. In the case of the lime-tree, individual 

 trees occur which produce pitchers annually and in large numbers, and 

 in the pitcher-plants they form a normal character for whole genera. 



Eeturning now to the abnormal structures of the flowers, two of 

 which are quoted by Masters as giving proof of the importance of terato- 

 logy for inquiries about the ancestral relations of plants : it is generally 

 assumed that the gamopetalous plants have been derived from the poly- 

 petalous groups. Authors may differ on the question whether this has 

 occurred once or several times along the main lines of the pedigree of the 

 plant kingdom. But the conception that free petals are the primitive 

 condition, and that gamopetalous or monopetalous flowers owe their 

 organization to a subsequent cohesion of the petals, may be regarded as 

 universally recognized. From this point of view, the dialysis of a 

 gamopetalous corolla has to be considered as a case of systematic atavism, 

 as a reversion of a species of the higher group to the more primitive con- 

 dition of older branches of the pedigree. 



The Khododendron (fig. 49) is one of the instances, and it occurs 

 among a group of gamopetalous plants which, curiously enough, include 

 some polypetalous species, such as those of Ledum and of Pyrola. 



Quite the reverse is the case with the perennial poppy. Some of its 

 varieties are producing, from time to time flowers in which the petals are 

 united by their margins, so as to form a large cup (fig. 50). Often the 

 petals are fused together in this way along their whole extent, but in 

 other flowers it is only more or less of their base in which they are so 



