330 SUMMARY AND CONCLUSIONS. 



become 90, 60, etc., by mimetic twinning. Such mimesie is very commonly 

 found in such polymorphous substances, whether the polymorphism be 

 due to physical isomerism or to chemical isomerism, and is generally desig- 

 nated as polysymmetry. The polysymmetric crystals of hemoglobins 

 appear in some cases to be the result of pseudosymmetry in the monoclinic 

 and orthorhombic forms that produces a tendency to a form of twinning 

 that favors such mimesie, and makes pseudo-tetragonal and pseudo- 

 hexagonal crystals; or it may be due to the pseudosymmetry being of such 

 a nature that externally applied force will cause a rearrangement of the 

 structure, with the development of a higher grade of symmetry. 



Many examples of this mimetic twinning or pseudosymmetry will be 

 found in the descriptions of the hemoglobins. It is especially common in 

 those groups which show normally the crystals of the hexagonal system, 

 as, for example, in the rodents. 



In the species of squirrel examined, the European red squirrel's crys- 

 tals of oxyhemoglobin are seen to be orthorhombic with a prism angle of 

 60 and hexagonal (or pseudohexagonal) with, of course, the same angle of 

 the prism. By twinning on the prism (really on the base as composition face, 

 but with the prism-base edge as the common direction in the crystals), three 

 of the orthorhombic crystals produce by their overlapping an essentially 

 uniaxial structure. As the orthorhombic crystals have the hexagonal 

 angle already, they naturally form hexagonal plates, which is the character- 

 istic tendency among the rodents in general. Probably all of the squirrel 

 crystals are thus twinned, when they show the characters of a uniaxial 

 substance, or when they are "hexagonal." 



A more striking example of such mimesie is seen in the case of the 

 ground-hog, Marmota monax. The y-oxyhemoglobin of the ground-hog 

 is monoclinic with a prism angle of 58 and an angle $ near 90. The 

 twinning is of the ordinary " horse- type, " fully described under horse and 

 mule. Three such crystals twinned on each other produce again a uniaxial 

 substance where the three overlap. The optical character is negative. The 

 axis of least elasticity a (=Bx a ) is within 10 of normal to the base. In the 

 averaging of the three axes to one in the pseudo-hexagonal twin, the other 

 axes b and c also average to a mean value, that of the ordinary ray of the 

 hexagonal crystal, and the optical character of the composite crystal, 

 considered as uniaxial, is also negative. Crystallizing de novo from the 

 solution appear large, perfectly formed, hexagonal plates of a-oxy hemo- 

 globin, which can not be distinguished by any power of the microscope 

 as not of homogeneous hexagonal texture. Their character is that of the 

 mimetic crystal that should develop from such a monoclinic crystal with an 

 angle of nearly 60 for its prism and an angle (3 nearly 90. Like the obvi- 

 ously composite crystals, the optical character is negative, but these show 

 a perfect uniaxial figure throughout, which the other only does in the center 

 of the overlapping group. 



Similar examples are seen in the case of the rats, when an obviously 

 composite hexagonal plate built up of orthorhombic individuals assumes 

 a hexagonal form by a similar sort of twinning; these are the hexagonal 



