Page:Mind (New Series) Volume 9.djvu/302

 288 NOTES. pressed with its working qualities. I prefer" [then follows a description of the writer's own apparatus]. A later account of this machine says : " It was made by Krille. I never used it for anything but rough demon- strations, and have since transformed it to other uses. It did not seem to me a good instrument." Instrument III. " The Krille pendulum worked satisfactorily for illus- trative purposes when carefully adjusted. I never had any confidence in it for scientific work, and never used it for systematic experiments. The same, however, is true of much of the apparatus used in our labora- tories." Instrument IV. " I have a complication pendulum, but made by Zimmermann [of Leipzig]. It is fairly satisfactory, except that the motion of the pointer is impeded when the mechanism for ringing the bell is raised, and that it is difficult to determine the objective time of the bell-stroke upon the scale." Here are five pendulums by Krille (if the reviewer's two be counted), and one by Zimmermann. All five of the former are subject, for one reason or another, to severe criticism. The Zimmermann machine is only 'fairly satisfactory'. The fact, then, seems to be that the design of the instrument, though meritorious in the first instance, is one that an improved technique and a wider range of problems render out-of-date ; that the workmanship, in most cases, is distinctly poor ; and that a desire to put a fairly cheap piece upon the market has further led to the use of poor materials. It would be as absurd to continue the charge of incompetency, in face of all these critics, as it would be to say that all German firms do less good work for exportation than they do for home consumption. THE KEVIBWBB. MR. MACCOLL'S QUESTION ON P. 144 OF MIND FOR JANUARY, 1900. In MIND, for January, 1900, Dr. MacColl invites expressions of opinion " as to whether the implication, If it is probable that A is certain, it in certain that A is probable, is always true ". By certain he means necessarily following from premisses or data (cf. "Symbolic Reasoning," MIND, Jan., 1900, pp. 79, 80). So the proposition hi question may be stated in this way : If it is probable that >S' is P necessarily follows from given data, then it necessarily follows from given data that S is P is probable. I think that thu inference here from Antecedent to Consequent does always follow, because what is pronounced probable from a "probability unsubjective point of view " must be so pronounced as a conclusion from given data or premisses. M is P ^ S ' M are P r k a bl e > * nen S is P is probably certain, but we can only say that - 1S jiT are probable, in the sense explained by Dr. MacColl, if we can produce premisses from which their probability necessarily follows. In the case which he gives on page 77 (MlND, Jan., 1900) if we take the figures numbered 1, 2, 3, as the possible alternatives, it is prob- able that Some E is A is certain. But this is only probable because Some E is A follows from two alternatives out of three ; so that the probability is certain i.e. follows necessarily from the given data. Thus, it will be observed, the Consequent follows from only part of the Antecedent, since probable = certainly probable. E. E. C. JONKS.