4 



Prof. Karl Pearson, and others. 



in 201 sprays from separate plants of Asperula odorata ; it was known 

 that these members are differentiated in their origin ; the homotyposis 

 was found much weakened. (17) and (18) The sori on the fronds 

 of 100 Hartstongue ferns and the lobes on the fronds of 100 

 plants of Ceterach were counted. "We were told that these charac- 

 ters are much affected by age of plant and environment of indi- 

 vidual; we found the homotyposis increased very sensibly beyond 

 the value obtained for trees. (19) The veins in the tunics of 200 

 examples of Allium cejpa were counted. (20) The seeds in the pods of 

 100 plants of Broom from Yorkshire were counted. In an Appendix 

 the homotyposis of the seeds in the pods of leguminous plants is dealt 

 with for a number of species. The general result is that homotypic 

 intensity is halved when we deal with a character associated with 

 fertilisation. 



W r e then considered two cases in which we knew the growth factors 

 to be very marked. Dr. E. Warren measured the length and breadth 

 ■of twenty-five leaves of 100 plants of common ivy (Hedera Helix) 

 and Dr. Lee and myself the length and breadth of ten gills of 107 

 Mushrooms (Agaricus campestris). The homotyposes of the leaf and of 

 the gill indices in these two cases were determined, and form series (21) 

 and (22). The homotypic correlation of the absolute lengths and 

 "breadths was also found in order to obtain some measure of the effect 

 of different stages of growth on homotyposis. Omitting the last series 

 of absolute measurements subject to growth, the mean value of the 

 twenty- two series gave the intensity of homotypic correlation as - 4570. 



(3.) A theory of fraternal hereditary resemblance is given on the basis 

 of the likeness of brothers being due to homotyposis in the characters 

 of spermatozoa and ova put forth by the same two individuals and 

 uniting for the zygotes whence the brothers arise. It is found that the 

 mean value of fraternal correlation ought to be equal to the mean in- 

 tensity of homotypic correlation. We have so far worked out nineteen 

 cases of fraternal correlation in the animal kingdom, and their mean 

 value = 0"4479, i.e., is sensibly equal to the intensity of homotyposis 

 in the vegetable kingdom. It is, therefore, very probable that heredity 

 is but a phase of homotyposis, and that the latter approximates to a 

 certain value throughout living forms. 



The theory involves a certain mean relation between direct and 

 cross homotyposis, i.e., that the homotypic correlation between char- 

 acters A and B in a pair of homotypes is the product of the direct 

 homotypic correlation of A and A (or B and B) and the organic corre- 

 lation of A and B in the individual. We had only the absolute 

 lengths and breadths of Ivy leaves and Mushroom gills to test this 

 proposition on, and the growth factor is here dominant. The results 

 ,do not show complete equality, but this is hardly to be wondered at 

 when we consider the extraneous influences at work. 



