754 



SCIENCE 



[Vol. LVI, No. 1461 



of the <7-,-'s that would be necessary to produce 

 his effect do not occur in ithe region of the vvorld 

 in which we live. 



I did not invent Einstein. I am not responsi- 

 ble for the theory of relativity or the deductions 

 made from it by physdeists and mathematicians. 

 It seems to me that Dr. Humphreys' criticism 

 ■siliould 'be directed to-ward them rather than 

 toward their humble interpreter. 



Edwin E. Slosson 

 Science Service, Washington 



ON THE FORMATION OF FAMILY NAMES 

 LIKE TINGID/E 



It is as'tonishing to observe how gi-eat a -dis- 

 plaj^ of erudition may he made in vain, the net 

 result being error. In i-eeent numbers of 

 Science Dr. Holland, Mr. A. C. Baker and I 

 have issued manifestoes on how to construct 

 family names based on &ird declension j-stems 

 not increasing in the genitive, and in ea<ih case 

 the argument has been vitiaited by a.t least one 

 mistake. However, each author has contributed 

 an item of truth, and it is now possible ito settle 

 the matter for good and aiU. 



As Dr. Holland says, the stem of the Laitin 

 word Tinge is undoubtedly Tingit-; 'but, as 

 Mr. Baker points out, Falbrieius did not adopt 

 this word, rather he introduced inito the neo- 

 Latin language the word Tingis, genitive 

 Tingis, stem Tingi.-. This brings us to my con- 

 tribution, i. e., -that Fabricius considered Tingis 

 ■'%i9 own. and indicated -whait its declension 

 should ibe" — 'perhaps a some-what misleading 

 sitatememt of the idea cleai"ly formulated by 

 Mr. Baker. My argumenit, 'however, had the 

 merit of reaching the right conolusion, namely, 

 tha(t Tingidce is the correct form for this family 

 name, and I have no ihesitation in diagnosing 

 as pathological the foiim Tingitidm in this par- 

 ticular case and Tingiidce or its like in all 

 similar cases. 



I 'have 'always bad a vague notion, founded 

 chiefly on unconscious observation, that in 

 forming patronymics from i-stems (noit in- 

 creasing in the genitive) ibhe final i of tlie stem 

 is to be dropped; and, indeed, who ever heard 

 of such ifcenns as Apiidse, Apihiidas, Feliida3 or 

 Caniidaj, until the publication of the last num- 



ber of the Proceedings of the Entomologicai 

 Society of Washington ?i To confirm or dis- 

 prove ithis 'belief and so to settle -the matter 

 beyond question, I lately addressed an appro- 

 priate quesition to Mr. Henry Pennypacker, 

 n'Ow of Harvard University and formerly 

 Greek teacher and headmaster of the Boston 

 Latin School. In reply I received the follow- 

 ing statement of the grammatical principle con- 

 cerned, as the joint opinion of my old teacher 

 and of Professor Clifford H. Moore, head of 

 the department of the classics at Harvard : 



Rules regarding .the fomnartion lof family names 

 which may bo described as patronymics are sub- 

 ject to modifioajtion not only in the interest of 

 convenience buit also of euphony, and dn spite of 

 the fact 'that the sitems of the nouns you mention 

 [Nabds, Apis, Tingiis, Coris, Aphis] in liatin end 

 in "i" and that the termination "idee" is con- 

 vantional in such cases there seems ■to be no doubt 

 tbalt the spelling with a single "i" carries uni- 

 versal authority and the penultimate " i " is short 

 in quantity. 



The authors •of the Intemiational Code, of 

 course, were fully conversant with this prin- 

 ciple and expected it to ibe applied in connec- 

 tion with Article 4, as it had been in the past. 

 Lest my acquaintance with the unexpressed 

 expectations of the members of the commission 

 ibe questioned, I should say that it is founded 

 on three considerations: (1) They were and are 

 educated men; (2) their own works contain no 

 such monstrosities as "Feliidffi" or "Anguiidas" ; 

 (3) authors and editors of standing throughout 

 the world have unanimously acted upon the 

 assum.ption which I have expressed above. 



There remains the widely tout not universally 

 accepted belief that priority should obtain in 

 family names, 'but the Code is not clear on 

 this point {i. e., Wh'at determines the type 

 genus of a family?) ; however this may be set- 

 tled in future, we arrive in the present instance 

 at the following conclusions: (1) that Tingidce 

 is nomenclaturally and philologically coiTect, as 

 Westwood was well aware when be proposed 

 the name in 1840; and (2) that it will not be 

 necessary to make the change in 'hundreds of 



1 The editor, Mr. A. C. Baker, substitutes the 

 term ' ' Aphiida ' ' far the term ' ' Aphididas ' ' used 

 by the au,thor of an ar.tdcle. 



