The Examination of Mr Tibb

I gave Mr Tibb the highest mark in comparative physiology that year, although he had attempted to cheat throughout the final examination and had devoted, by his own subsequent estimate, eleven weeks to preparing the means. This decision occasioned some correspondence. Professor Harlowe considered it destructive of academic discipline, and the Registrar asked whether the Royal Collegium now proposed to recognise fraud as a branch of natural philosophy. I answered both gentlemen at sufficient length to discourage a second enquiry. Since Harlowe has nevertheless continued to describe the affair at dinners from which I am absent, I shall set down what Tibb actually contrived and why his distinction was earned.

Tibb, a Gremlin, occupied the front bench because the benches behind it afforded him an excellent view of his classmates’ coat-buttons. He had a wooden platform for his feet, which I had ordered after finding that the porter expected him to sit through lectures with his legs dangling. His written work was uneven: he could trace the circulation through a mammalian heart accurately, then lose half his marks by declaring that a valve closed “when required”. I wrote beside this that the valve had neither a timetable nor a conscience, and requested an account of the pressures acting upon it. He came to my room to argue that the pressures were implied. I made him calculate them. Thereafter he attended and argued with equal regularity, submitting work whose increasing correctness was accompanied by an unfortunate decline in legibility.

The final examination joined written questions to a circuit of demonstration stations. At each station stood an enclosed preparation or apparatus, and the candidate had to identify its action, predict the consequence of some alteration, and explain the prediction. The cases covered circulation, respiration and the expenditure of muscular energy. Students were allowed an ordinary timepiece, since several exercises required them to count an interval. Tibb petitioned to bring a larger timing instrument, saying that the small hands of a watch were difficult to read while observing a moving preparation. This was reasonable. I approved it without inspection, a concession he afterwards identified as essential to his success. I had imagined that a machine which required a carrying handle would be an inconvenient vehicle for concealed assistance; its inconvenience was among its principal recommendations.

He arrived with a brass case about the size of a substantial dictionary and set it beside his ink. Its dial carried a broad seconds-hand. Below it were two winding keys, a row of little sliders, and a shutter through which appeared such innocent reminders as “count full interval” and “repeat observation”. Tibb explained that these were aids to the timing procedure. I told him that any candidate who needed a machine to remind him to repeat an observation had neglected my lectures. He bowed with an expression I mistook for contrition. I should have remembered how readily students agree with a professor who is standing between them and an unexamined brass box.

For the first hour I noticed only that Tibb took unusually careful measurements. While other candidates guessed at the narrowing of a tube, he borrowed the graduated wedge provided at the station and read it twice. He counted the strokes of a pump over a full minute, entered something upon his instrument, wound the left-hand key and waited. At the fourth station, however, his seconds-hand stopped while a faint succession of clicks continued within the case. He then opened the shutter and wrote rapidly. I crossed the room and asked him to lift his hands from the desk. He obeyed, though he first attempted to close the shutter with his wrist. The exposed strip read: “Increased resistance beyond the pump; chamber pressure rises before delivered volume recovers. Examine relief passage.”

I removed the case. Tibb said that this was a timing instruction. I read it aloud and asked him which part told the time. He considered the question longer than it deserved, then requested permission to explain privately. I allowed the other candidates their full remaining interval, placed the instrument beneath my chair, and required Tibb to finish upon blank paper. He completed two answers and spent the rest of the examination watching my feet. Afterwards I summoned the demonstrator, Miss Cresset, whose presence proved valuable both technically and because Tibb later claimed that I had threatened to open the machine with a coal hammer. I had merely enquired whether one was available.

The apparatus contained no copied lecture sheets. Tibb had obtained rejected brass strip from an instrument-maker and fashioned it into a set of perforated bands, each carrying abbreviated conclusions. The sliders admitted observations: whether the working chamber enlarged or contracted, whether flow increased, where a constriction had been introduced, and whether the circulating fluid encountered an exchange surface before returning. Turning the key drove feelers against the bands. Certain combinations released a sentence at the shutter; others locked the mechanism and displayed “insufficient observation”. A second train governed exceptions. I examined the device for several minutes before speaking, which encouraged Tibb to suggest that it was substantially a calculating aid. I reminded him that calculating aids permitted in my examination did not ordinarily compose the candidate’s explanations.

Its concealment was more elaborate than its exterior suggested. Opening the lid disengaged the answer train and left the timing mechanism running. Tibb had expected a cursory inspection and had made the visible works genuinely serviceable. The sentence bands could be withdrawn through the winding aperture with a hooked wire; he carried that wire inside a penholder. He had even arranged for the shutter to show reminders whenever the case tilted towards an observer. The stoppage that exposed him arose because the clock and the answer train shared a spring, and he had entered a combination which demanded an unusually long search. Here I recognised his familiar weakness: he had perfected several ingenious protections against being discovered and economised upon the one component necessary to make them all work.

I was prepared to record a failure when Miss Cresset asked how the machine dealt with the sixth station. This concerned a respiratory preparation whose exchange membrane had been thickened without changing its area. Tibb showed us the relevant setting, then volunteered that the answer depended upon whether the fluid passed the membrane once or repeatedly during the measured interval. I had altered that feature shortly before the examination, and he could have had no advance knowledge of the alteration. When I asked how his machine had provided for it, he removed a band and laid it upon the desk. Its branches distinguished changes in exchange area from changes in the difference driving the exchange, and both from changes in the time available. In the margin, scratched so finely that I required a lens, was a correction to an example in my own lecture notes.

I objected to the correction. Tibb fetched his notebook, turned to an experiment he had repeated after class, and objected to my objection. I had assumed that the circulating fluid remained equally receptive along the whole membrane; in the apparatus described, it did not. My approximation was suitable for the beginning of the process and increasingly poor thereafter. He had measured the discrepancy because it caused his mechanism to select contradictory statements. We repeated the calculation together. He was right. I corrected the lecture notes that evening, and have retained his figures. Harlowe has suggested that gratitude for this service softened my judgement. He appears to suppose that I value a correction less than the comfort of avoiding one, a mistake readily explained by acquaintance with his publications.

I then began an examination of my own. I gave Tibb a circulating system with a leak before the pump and asked what his machine would advise. He set the sliders, read the result, frowned and withdrew it. The device assumed a rigid inlet; a collapsible one could narrow under the altered pressure and reduce delivery still further. I asked him to account for the difference at the board. He did so, drawing the successive conditions and marking where his information ceased to support a numerical answer. I changed the fluid’s thickness. He amended the argument. Miss Cresset introduced a valve which closed incompletely, and he explained why a satisfactory stroke count might conceal a considerable loss of useful flow. We continued until the porter came to extinguish the corridor lamps.

The eleven weeks emerged during this discussion. Tibb had begun by trying to predict my questions from previous papers. Finding that I changed the particulars, he had attempted to classify every question I could reasonably ask. Each classification required him to decide which observations mattered, and every exception demanded another branch that threatened to make the brass bands too long. To fit the case, he had repeatedly reduced explanations to their essential dependencies. He had borrowed apparatus, persuaded a machinist to let him use a small punch, and spent evenings testing whether one conclusion could safely serve several different preparations. He had also recruited two classmates to enter misleading observations, so that he could discover where the device gave a confident answer upon inadequate evidence. They believed they were helping him construct a teaching instrument. On this point I questioned them separately.

I asked why he had undertaken such labour to avoid an examination for which that labour had thoroughly prepared him. He said that he trusted mechanisms: when obliged to write under supervision, he found himself remembering my objections before he could remember his own argument, and revised sound answers into confusion. The instrument would commit itself when he could not. This explained the scheme without excusing it, and I told him so. He had deceived his classmates as well as his examiners, and had sought an advantage denied to candidates who obeyed the rules. I also told him that he had been remarkably foolish in supposing that an examiner would be less alarming after discovering a concealed sentence-engine. He conceded this with more readiness than he had shown over the valve.

The examination regulations allowed a supplementary oral assessment where a written performance could not be accepted. I rejected his original script and required him to return the following morning, without the apparatus. Miss Cresset and I examined him for two hours upon fresh problems, including matters his bands did not cover. His explanations were the best offered by any candidate that year. Accordingly, I gave him the highest mark. The disciplinary board censured him, and he apologised to the classmates whose assistance he had obtained under false pretences. I subsequently asked him to rebuild the instrument openly for teaching. The revised model occupied twice the space, exposed its assumptions beside each answer, and required students to explain why its conclusions followed. Tibb wanted my name engraved upon it; I insisted upon his own, together with a plate recording that the mechanism supplied no warrant for the correctness of the observations entered. He complained that the wording was severe. I invited him to propose a defensible amendment, and we spent a thoroughly agreeable afternoon disputing it.