304 



NA TURE 



[July 28, 1892 



several years from the birth of the last phthisical child. Here 

 we should be driven to assume, not in the case of the mother 

 alone, but in each of the several children, a long latent period, 

 during which the parasites, though present in the tissues, made 

 no sign. Such an assumption presents great difficulties. Again, 

 the direct transmission of tuberculosis from a mother to her 

 foetus is admittedly rare, whereas on the supposed hypothesis 

 we should expect to find it common.^ 



But if it is not the parasite that is transmitted, what is 

 transmitted? We are driven back on the "other side" of the 

 pathology of phthisis. But if we suppose that the transmission 

 is not one of a parasite, but of a "diathesis," or " predisposi- 

 tion," then we desert the only standpoint from which there is 

 any chance of proving that the disease was acquired in the 

 sense under discussion. For what reasonable ground could we 

 have for restricting this "predisposition" to the somatic cells 

 alone, to the exclusion of the reproductive cells ? 



On the hypothesis that the thing transmitted is a "predis- 

 position," we can, as in gout, explain the element of progressive 

 heredity in phthisis. For, the admission of a morbid change 

 once made, the difficulty is not so much to explain its progres- 

 sion as its arrest. In certain consumptive families we have in 

 the limits of a single generation this morbid progress going on 

 under our very eyes. It is the rule to find in such families, 

 where several brothers and sisters are attacked, the younger fall 

 victims at an earlier age than the elder, showing in this way 

 their increasing liability. The explanation is probably identical 

 with the one suggested in gout. The entire organism of the 

 parent becomes more and more phihisically disposed— somatic 

 and reproductive cells alike. The later the separation of the 

 latter occurs, the more likely will they be to manife>t phthisis. 



The same line of argument is applicable to the facts of 

 "short sight."- Short-sightedness is certainly hereditary — it 

 runs in families — but that does not prove that we have in it an 

 example of the transmission of acquired characters. For in the 

 first place it would be very difficult to prove that the short sight 

 was in the first instance acquired in the sense under discussion. 

 While the progressiveness of the morbid character — which 

 seems to support the theory — can be as well explained without it. 

 For if there is no proof that the morbid character — the faulty 

 build of the eye — is itself progressive, there is good reason to 

 suppose that the habits of close attention which minister to the 

 defect are so. In one generation we find a man simply tasking 

 his eyes ; his son works with a simple microscope ; his grandson 

 with an improved microscope. 



I pass on to consider another group of pathological facts, of 

 the highest importance and interest — new growths. The element 

 of heredity doubtless obtains here as in the case of gout and 

 phthisis. Thus the statistics of Sir J. Paget in this island, and 

 those of Velpeau on the Continent, agree in showing that 

 heredity can be traced in about one-third of the entire cases of 

 cancer. 3 And among the benign tumours, as they are called, 

 warts and exostoses are hereditary. Further, there is in some 

 cases evidence of progressive heredity, the irregularity appearing 

 in the children at an earlier age than it appeared previously in 

 the parent. And we have here what might look at first sight 

 more like a real transmission of acquired characters than any- 

 thing we have yet dealt with. No one questions that something 

 is transmitted. The theory of the local origin of the new growths 

 is gaining ground everywhere, and might appear to carry the 

 inference that they are acquired, and that no constitutional 

 element is involved in them. Here, however, we must be on 

 our guard against the fallaciousness of words. If by consti- 

 tutional we mean something pervading the entire organism — a 

 taint in the blood, and so forth — then there is little or no 

 evidence to warrant our calling new growths constitutional. 

 But if we mean, on the other hand, something which was 

 represented in the original germ — an error in the original plan, 

 not a supervening flaw — then theie is nothing to encourage us 

 in denying, and a good deal to warrant our asserting, their con- 

 stitutional origin. However, such an admission is not necessary 

 to our present purpose. Let us assume that they are acquired 

 in the sense in which a scar is acquired. Is it a fact that what 

 is acquired is transmitted ? If so, we should look for identity in 

 position and histological character in the thing transmitted. 

 But on the whole neither of these conditions is fulfilled. Cer- 



1 See Fortschritte de7- I\Iedicm, Bd. iii., iSSs, p. 19S ; b.acilli found in 

 lungs of foetal calf, set. 3 months, whose mother was tuberculous. 



2 Ziegler and Nauwerck, " Path.," vol. i. pp. 393-94. 



3 Erichsen, " Surgery," seventh edition, p. 7S7. 



tainly they are not, in the case of cancer, as the analysis by Mr. 

 Morrant Baker ^ of 103 of Sir J. Paget's cases clearly shows. 

 The distribution of the cancers proper shows a variation within 

 certain limits. There is a strong predilection for certain sites, 

 but these sites are sufficiently numerous. Now, it often happens 

 that, where several children inherit cancer from a parent, the 

 growth appears in each case in a different site. Nor are the 

 precise histological characters of the growth at all faithfully 

 preserved in the course of transmission ; while it has been often 

 observed that on the bodies of cancerous people innocent 

 growths exist as well." So that the inheritance does not appear 

 to be a liability to a peculiar modification in a certain part, but 

 a tendency to one or more of a group of modifications in one of 

 many possible sites. 



Once more we find ourselves driven to a choice between two 

 alternatives, either of which excludes the transmission of acquired 

 characters. P'or if new growths are really acquired characters, 

 then it is not exactly what is acquired that is transmitted, but 

 something broader than it. If, on the other hand, they are only 

 acquired in a more general sense, they fall outside the limits of 

 Weismann's sense of the term " acquired character." 



(3) There remain for our consideration the third, and, in one 

 sense, the most important, group of pathological data — those 

 which answer to the qualifications of acquired characters in 

 Weismann's stricter usage of the term. Here, if anywhere, 

 would be the ground in pathology to select for proving the 

 theory of the transmission of acquired characters ; but it must 

 be confessed that this is just the region in which that theory 

 receives the least support. This group of pathological facts 

 embraces a number of accidental lesions, such as scars and 

 mutilations, which are certainly acquired in the strictest sense 

 of the word. But the evidence for the theory seems strong only 

 in the dubious cases, weak in the unexceptionable ones. We have 

 examples of mutilations practised for many centuries by entire 

 races, without being transmitted in a single instance. Nor is it 

 the experience of surgeons that scars and mutilations which are 

 the results of operations are ever transmitted. On the othei- 

 hand, we have histories of tailless cats and hornless cows. But 

 here everything turns upon the comparative certainty with which 

 we can prove that the initial lesion was really in the first in- 

 stance acquired. Have we here to do with an accidental lesion 

 or a deformity ? A clcser investigation has, in many instance.*, 

 rendered the latter the more probable explanation of the two. 

 For example, in the case of the tailless cats, closer research made 

 it appear that the irregularity involved an abnormality affecting 

 many of the lower vertebrae. In other cases, the abnormality 

 in the child was so little like that in the parent, as to suggest 

 that it was a merely accidental coincidence of two different 

 lesions in one site.^ 



If we turn to the results of experimental research, we are con- 

 fronted by more than one remarkable series of experiments, 

 upon the bearing of which it is impossible as yet to pronounce 

 decisively. The most notable work done in this direction is, 

 perhaps, a series of experiments upon guinea-pigs, undertaken 

 by Brown-Sequard, and repeated by Westphal.^ They produced 

 epilepsy in a number of these animals by various methods 

 — section of the cord, section of different nerves, &c. — and ob- 

 served subsequently that certain of the offspring were epileptic 

 too. 



But there are several reasons which prevent our accepting 

 these results as decisive. The records of the experiments are 

 said not to be very perfect. Then it is not contended that 

 epilepsy was uniformly transmitted. What happened was that 

 each member of the offspring presented some morbid symptom — 

 usually some nervous trait, such as epilepsy or paralysis. So 

 that the result of Brov\n-Sequard's experiments would rather 

 seem to be this. By producing one morbid trait in the parents, 

 he set up a liability to one of several in the offspring. By pro- 

 ducing a single character, he set up a tendency. All this is of 

 extreme importance, and it may well be that the future has much 

 that is interesting to reveal in this direction. But, meanwhile, 



1 See "St. Bartholomew's Hospital Reports," 1866. 



2 Observation of Mr. J. Hutchinson, quoted in Fagge's " Medicine, 

 vol. i. p. 106. 



3 For a number of other instances, see Weismann's essay on Ihe bup- 

 posed Transmission of Mutilations," passim. 



4 See Krown-S6quard. "Researches on Ejiilep-sy," Boston. 1857 ; Papers 

 \-nJo7n7ialde Physiologic de V Homme, torn. i. and iii., 1858. i860 ; Archives 

 de i>hysiologicnorimjle etpathologiguc, tom. i.-Lv.. 1868-1872. Ziegler and 

 Nauwerck, vol. i. p. 390. See also Weismann on Brown-Sequard, pp 81, 

 310, 313 ; translation, edited by Poulton. 



NO. 1 187, VOL. 46] 



