192 



SCIENCE 



[N. S. Vol. XLV. No. 1156 



under the name of genus- or species-specificity 

 are Mendelian in character. It is obvious that 

 a definite answer to this question would be of 

 fundamental importance for the problem of 

 evolution. If species-specificity is not a Men- 

 delian character, we are confronted with the 

 possibility that Mendelian mutations may not 

 have been the only essential factor in evolu- 

 tion. 



The phenomena of species-specificity are, as 

 far as we know at present, exclusively deter- 

 mined by the proteins. Phenomena of cyto- 

 lysis by foreign blood or extracts of foreign 

 tissues, the precipitin and the anaphylaxis re- 

 actions can apparently not be produced by any 

 other constituent of an organism than the pro- 

 teins ; and the two or three exceptions reported 

 to this general rule may have been due to im- 

 purities in the substances used for the experi- 

 ments. 



A decision of the question of heredity men- 

 tioned might be possible by comparing the spe- 

 cies-specificity of a hybrid with that of the two 

 parent forms. If it could be shown that the 

 species-specificity of a F, hybrid is identical 

 with that of only one of the two parents, no 

 matter whether this parent is the paternal or 

 maternal species, we might consider this an 

 indication that species-specificity is Mendel- 

 ian; if the species-specificity of a Fj hybrid, 

 however, is always identical with that of the 

 maternal form, no matter from which of the 

 two parent forms the mother is selected, it 

 might indicate that the cytoplasm of the egg 

 determines the inheritance of the species-spe- 

 cificity. 



Experiments of this kind meet with the diffi- 

 culty that only closely related species can be 

 crossed successfully and in closely related spe- 

 cies the differences in species-specificity are 

 generally too uncertain to permit a definite 

 conclusion. In the splendid work of Reichert 

 and Brown on the " Differentiation and Spe- 

 cificity of Corresponding Proteins and other 

 Vital Substances in Eelation to Biological 

 Classification and Organic Evolution," ^ it 

 has been shown that the corresponding hemo- 

 globins of different species are not identical, 



2 Carnegie Institution Publication, No. 116. 



and " that their peculiarities are of positive 

 generic specificity and even much more sensi- 

 tive in their differentiation than the precipitin 

 test." In their book they describe the hemo- 

 globin crystals of the horse and the mule, but 

 not those of the donkey. It seemed of interest 

 to make the series complete in order to find 

 out whether or not the hemoglobins of the 

 mule resemble more closely the maternal or 

 paternal form. A decisive result in favor of 

 a Mendelian origin of the specificity could only 

 be had if the hemoglobin crystals of the Fj hy- 

 brid were identical with those of only one of 

 the two parent forms, otherwise the result 

 would decide neither for nor against a Men- 

 delian inheritance of species-specificity. 



The writer obtained donkey blood, the hemo- 

 globin crystals of which were prepared and 

 analyzed by Professor A. P. Brown, of the de- 

 partment of rnineralogy at the University of 

 Pennsylvania, who was kind enough to com- 

 municate his results to me in a letter which 

 with his permission I take the liberty of pub- 

 lishing here. 



May 4, 1916 



Dear Dr. Loei : I suppose you have come to the 

 conclusion that I have forgotten all about your 

 samples of the donkey blood, containing no oxa- 

 late, which you so kindly sent me; but I have been 

 working upon them as my time permitted and I 

 think that I can now venture to state that in the 

 orthorhombic [or what I have called the " a-oxy- 

 hemoglobin"] constant differences may be ob- 

 served; and these indicate that in this substance 

 the blood of the donkey more closely resembles the 

 blood of the horse than it does that of the mule. 

 I place more weight upon the results obtained from 

 the orthorhoviiic crystals than upon those deduced 

 from the mo7wclinio crystals for the reason that 

 the monoclinie crystals obtained from all three 

 bloods show a strong tendency to tmu and these 

 twins are what we call "mimetic twins." The 

 name "mimetic" is applied to them because they 

 mimic or imitate a higher grade of symmetry than 

 they really possess. For instance, these mono- 

 clinic crystals approach in their angles those of 

 the hexagonal system and are indeed what we call 

 pseudo-hexagonal. For these somewhat plastic 

 crystals, by the way they twin, average their asym- 

 metries (i. e., their departures from pseudo-hexa- 

 gonal symmetry) until they become, in their angles, 



