MISCELLANEOUS: STIMULI UNDERLYING TUMOR-FORMATION 511 



attributable to slight variations in the direction or force of the 

 stimuli and to the diverse cell-inheritances, each and every tissue 

 responding according to its own specific nature. All tumors 

 begin, so far as we know, in injured places 1 and, fundamentally, 

 I believe all may be regarded as excessive and continually modi- 

 fied wound-repair reactions. 



In this chapter I shall deal with the secondary causes of 

 tumors and shall endeavor to present my ideas as briefly as 

 may be, premising that they are based on experiments and 

 that where they pass beyond experiment into the field of hypo- 

 thesis, no one need be led astray, if he keeps my title in mind. 

 The best of any iconoclastic writing in science is not so much 

 the new facts it has to offer as the changed outlook it gives, 

 which new perspective often leads to renewed important experi- 

 ments and to general discussion by many workers. I may claim 

 to have contributed, at least, this much toward the elucidation 

 of the complex and important problems involved in the origin 

 of tumors. 



I may state at the outset that my conception of tumor forma- 

 tion involves a loss of water and a change of chemical reaction 

 in the cells which are to become tumor cells. This change which 

 is toward cell-sap concentration and increased acidity must 

 occur, I believe, to give the necessary stimulus to tumor forma- 

 tion. The stimulus may be long-continued or fleeting and may be 

 brought about, as we shall see, in various ways. I will develop 

 the subject as I proceed, and only add here at the beginning 

 two other hypothetical postulates, first, that hyperplasias appear 

 to me to represent a reponse of cells to oxygen-hunger or 

 semi-asphyxiation, and, second, that the type of cell-response 

 in the tumor, that is, whether a simple cell-enlargement with 

 mitotic or amitotic nuclear multiplication (a giant-cell), or 

 a full karyokinetic nuclear division with a corresponding 

 hyperplasia, appears to depend on whether the partial proto- 

 plasmic cell-paralysis involves both nucleus and hyaloplasm, 

 or is confined to the latter, leaving the nucleus free to divide by 



1 Of 850 breast cancers studied by MacCarty in the Mayo Laboratories every 

 one showed evidences of having been preceded by inflammatory injuries (chronic 

 mastitis). 



