168 R. H. MOLE 



mfller: What's the difference between that and a "latent period"? 



MOLE: Well, I don't normally use the plirase "latent period" which has all sorts of 



meanings to different biologists. I have said that there is a certain probability of inducing 



the tumour which increases from zero time. And I have just subtracted from the survival 



time the time required for the growth of the tumour assessing that on the average this is 



the same for all tumours. And so I tried to avoid this idea of latency. 



muller: May I go back to the question which I asked originally? If you make a suitable 



assumption, neglecting its plausibihty, can you make these data fit either a one hit, a two 



hit or a three hit hypothesis? Because if so, the reliability of any of these hypotheses 



seems open to question. 



MOLE: Well it depends on what you mean by differences in assumptions. Brues had to 



assume that tumour incidence depended on the number of radioactive events, and also 



on a latent period, unspecified as to meaning and varying with dose. This further 



quantitative assumption was only made in order to draw straight lines. I thought I had 



reduced the number of assumptions by at least one anyway! 



BERENBLUM: The fact that it is difficult to define latent period doesn't mean that it 



doesn't exist. The point is, if you adopt another term, and then assume that it is a fixed 



period and deduct it, in a sense you may be introducing an error In other words to 



assume that this latent period of whatever you call it, is a fixed time may be Vrong, it 



may be proportional to the total time in which case would it not induce errors in your 



calculations? 



MOLE: Oh well, all I was saying was that if you don't assume a fixed value for it, the data 



will not fit the proper lines. That's all I'm saying. 



UPTON: I would like to speak again about the question of the Japanese data and parallel 



curves. To me this is an extremely revealing situation, in that, I am told that the more 



recent data from Japan on tumour incidence are consistent even though the numbers are 



stni very small. But what experimental evidences we have are consistent with the idea 



that the log of the death rate for tumours plotted against age is displaced by giving doses 



of radiation. From the raw data it doesn't seem to matter how old the individual is, you 



seem to get about the same degree of displacement. As Dr. Mole emphasized since this is 



a log-scale, doubling the tumour incidence at an early age represents a far smaller net 



number of tumours induced, than doubling the tumour incidence at a later age. In this 



sense radiation is m fact adding something aldn to time, if you simply look at actual data. 



ALEXANDER: Would the situation be, Dr. Mole, that with regard to the bone-tumour 



data, they can be fitted to a one hit law, when one assumes a latent period which is 



proportional to dose, and can be fitted to a two hit law if j^ou assume a latent period 



independent of dose, and therefore, we can take our choice as to wliich we think is more 



probable — the latent period mdependent of dose or latent period dependent on dose. So 



really you haven't thrown one assumption away, we have stiU got two alternatives, and 



if we once know about the latent period, then we can decide whether we want a one hit 



or a two hit event. 



MOLE: If you like to think of latent period as being anything more than a mystical 



assumption. 



mayneord: I have a suspicion that something of the same kind occurs if you look at the 



normal incidence. Dr. Turner and I have recently been very interested in the question 



of the incidence of cancer of the stomach in relative to radioactivity. In fact, we find very 



little, but this has led us to plot, let us say the incidence of carcinoma of the stomach 



against age. And you get varying degrees below, I think about a 5-8 or 6th power. 



Incidentally, I still don't understand how you fit a square law to that! What does 



iiitrigue me, and I wish Prof. Berenblum would tell me about it, is that if you take the 



