Living in Qing China, you’d probably smoke opium. You’d thereby embroil yourself, one way or another, in the opium wars of 1840 or so, in which the Brits force-sold opium to the Chinese to rustle up the cash needed to buy tea from them. And so you’d find yourself at the very start of the ‘century of humiliation’ that figures heavily in the historiography Chinese leaders up to today rely on.
A few decades later and across in Paris, you’d take laudanum (alcohol with opium) and weed, and praise ‘artificial paradises’ as the urban world came to life around you.
At turn of the 19th century, in the US and elsewhere, you’d go to the pharmacy with a cough or the shits or neurasthenia and get dope and coke and weed.
In 1940 you’d instead knock yourself out with barbiturates, a sedative easy to overdose on. Nazis pump out tons and tons of meth; it is contained in soldiers’ rations and consumed in a wild cocktail by Hitler.
In the 50s you might pop tranquiliser Miltown, which actors joke about on late night tv. In the 70s, of course, you’d smoke weed.
In the 90s, in Russia, you’d drink and trade and bribe officials with vodka. If you ran out, you’d substitute with samogon (homebrew) or eau de toilette or shoe polish. Same decade, in the US, if you’re black, you smoke weed and get imprisoned. If you’re white, you smoke weed and don’t go to prison.
And now? Now you have a weed vape and set your alarm for 5am Adderall. You probably drink, although could be “LA sober”; you might vape tobacco too. If you step outside the law you get good (and cheaper) weed, cut coke and “dirty m30s”, pills bearing a slight resemblance to Oxycontin but that you all know is fentanyl or worse.
History shows that the drugs available to us is not a given like the phases of the moon or the need for sleep: it varies from place to place. Moreover, there appears to be some causal story about why we find this or that drug in a given society: because the British needed something to trade for Chinese tea or the world war spurred research into chemistry or the Russian state often relied on vodka taxes to survive or the US policing system is racist.
My big picture aim is to classify those causal facts, the better to understand why it is we have the drugs we do, and how we could have better drugs, drugs that could lead better to human flourishing. (And I’m very open to the possibility that the best drugs we could have are none at all!)
But that is a very big picture aim, the sort of thing one can get lost trying to investigate. Accordingly, in this post, I will narrow things down massively, trying to pose and answer one relatively precise question. I think in answering this some of the big-picture facts will become evident.
Here is the question:
In 5 years, will functionally-amphetamine-like stimulant use in the US be lower, higher, or roughly the same as it is now?
Very unfortunately, this is a bad question, but I can’t think how to make it better. It’s bad because ‘roughly the same’ is underspecified: are we talking absolute numbers, population-scaled numbers, rates of change, or what? It’s bad too because it’s hard to answer. I spoke intentionally of ‘use’ so as to include both prescribed and unprescribed use. But while we can at least ballpark the former with prescription data, the latter is going to be hard to find good data for, and even if we could aggregating the both would be difficult. Finally, ‘functionally amphetamine-like’ is vague but unavoidably so (does it include modafinil, to take just one problem case?) We live in an era when new substances are discovered and used rapidly; it’s highly possible that something with amphetamine’s effects but a sufficiently different chemical structure be discovered and become mainstream.
Despite its flaws, this is our question. I call this a ‘choose-your-own-conclusion’ essay because my aim is more to present a set of considerations that I think are somewhat relevant to answering the question than strongly to defend an answer (though I will defend an answer). Hopefully some will find it fun to consider the strength of the bits of evidence I consider and the crude model I construct and come to their own conclusion (I would have made a prediction market to associate with the essay but didn’t both from laziness and from disliking my question). So ponder the question now, and then see if your initial guess changes having read the below.
ADHD and ADHD Meds
There are two types of ADHD. A sufferer of the first type (I quote Wikipedia):
Frequently overlooks details or makes careless mistakes
Often has difficulty maintaining focus on one task or play activity
Often appears not to be listening when spoken to, including when there is no obvious distraction
Frequently does not finish following instructions, failing to complete tasks
Often struggles to organise tasks and activities, to meet deadlines, and to keep belongings in order
Is frequently reluctant to engage in tasks which require sustained attention
Frequently loses items required for tasks and activities
Is frequently easily distracted by extraneous stimuli, including thoughts in adults and older teenagers
Often forgets daily activities, or is forgetful while completing them.
The second is typified by (a person who):
Excessive motor activity; struggles to sit still, often leaving their seat; prefers to run about; in younger children, will fidget when attempting to sit still; in adolescents and adults, a sense of physical restlessness or discomfort with being quiet and still. Talks too much; struggles to quietly engage in activities. Blurts out answers or comments; struggles to wait their turn in conversation, games, or activities; will interrupt or intrude on conversations or games. A lack of forethought or consideration of consequences when making decisions or taking action, instead tending to act immediately (e.g., physically dangerous behaviours including reckless driving; impulsive decisions).
Amphetamines are a type of drug that stimulate the brain. Much is unknown about how they work, although at the largest scale we know that if you take them you’ll feel good and energized (as well perhaps, as anxious and on edge), and at the smallest scale we know that they will interfere with brain chemicals called dopamine and noradrenalin and in certain ways cause the amount of those chemicals to increase in the brain. We know that increased (or differently behaving) dopamine in the brain correlates with certain positive mental states like (very roughly) satisfaction or reward, and that noradrenalin sort of speeds up the communication between the neurons in your brain, and knowing two causations (eat pill makes feel good, eat pill change brain) and two correlations (more dopamine feel good, more noradrenalin quicker brain) we have a sort of story of what’s going on.
ADHD as a malady, and with it amphetamines as a treatment, are relatively recent. People had been noticing something roughly like that for a while, but we can start the story with its medicalization. As with much in the history of psychoactive drugs it was a mix of mistakes and actions that would never pass a review board. Someone tried to make nasal decongestants; one such product made him feel good. In the absence of legislation (this was the 30s), the drug company sent out samples to various doctors, one of whom was bold enough to try it on some hyperactive kids (to reduce their headaches!) On giving it to them, he was amazed to discover them pacified. (Source for all this: Schwarz ADHD Nation ch1).
Many people take amphetamines for ADHD, and the number is increasing a lot: in the period 2017–2021, prescriptions increased about a third up to about 40 million. According to one study, adult ADHD’s prevalence more than doubled between 2007 and 2016; and adults are now diagnosed at 4 times the rate of children.
This might sound troublesome, but arguably isn’t. In the UK, for example, over a roughly similar period of time there was similarly just over a 33% increase in anti-depressant prescriptions, which is noteworthy because antidepressants have no recreational use (see; in a fuller treatment, more attention would be given to these trends in prescribing of other drugs).
You might have noticed this, especially if you happened to read any of the stories about the often-called ‘mysterious’ shortage in ADHD drugs. One theory is that in the pandemic people increasingly resorted to telehealth, and increasingly got prescriptions; but, amphetamine is a class 2 scheduled substance (“high potential for abuse which may lead to severe psychological or physical dependence”) and so even its medical use requires a bunch of regulation and special treatment that can’t be scaled up super quickly. At least, that’s one theory, but the main reason I bring this up is to talk about why it’s schedule 2.
A drug is schedule 2 if there’s some concerns about it. Now, the scheduling system is in general idiotic (weed is 1 and fentanyl 2), but even a stopped clock is sometimes right, and there are indeed reasons to be worried about amphetamine (this isn’t to impugn it: almost all drugs have problems, none more so than alcohol and tobacco.) Amphetamines are bad on the heart leading to hyper- or hypo-tension and an increased heart rate. Very excessive use can lead to psychosis, and you’ve surely encountered many stories of the life-ruining effects of stimulant use in the rust belt or on SBF’s island. (Most of the stories one reads about concern methamepthatmine rather than amphetamine, and often involve very large doses of impure drugs via more dangerous routes of administration like snorting and injecting, and often in the context of a life led at the margins of social life; it’s unclear if meth horror stories should overly concern the average Adderall user.)
The Wikipedia notes that dangerous side effects are not a concern at therapeutic as opposed to recreational doses and that although amphetamines are very addictive, they are not so at those therapeutic doses. In a sense, this claim will be the crucial one. As we’ll see, there’s substantial precedent for drugs that were claimed to be non-addictive only to turn out to be very addictive.
The Future of Amphetamines
That, then, is a very quick overview of some of the issues surrounding amphetamines, which can be summarised as saying that it’s an increasingly popular class of drugs that help people focus and feel good but that carries with it potential risks of addiction and other side-effects.
Let’s assume that if you want amphetamines in the US today, you can get them. Is that good? Will that change? To answer that I’m going to sketch and evaluate four paths, which I will call eradiction, whereby the dangers of amphetamine will become apparent, it will be prescribed less, and its users won’t have recourse to illegal alternatives (the historical analogue being Miltown and benzodiazepines of the 70s); legalization, whereby current trends in use is a sort of interregnum between illegality and wide recreational legal use, as we are seeing with weed; criminalization, whereby ill effects become undeniable but addicted users have recourse to alternatives, as we see with oxycodone and fentanyl; and finally a least resistance path whereby the role of amphetamine remains the same.
The increase in amphetamine prescriptions is far from unprecedented. But the drugs whose use increases varies with time and place. If you were living in the 50s, for example, instead of reading stories about Adderall shortages and ADHD, you’d probably be hearing about the so-called tranquiliser ‘Miltown’. (The word, and indeed the concept, of ‘tranquiliser’ was new: prior, one had ‘sedatives’. Miltown was a drug to remove anxiety, not to knock you out.)
This was, in its time, an object of pop culture; if Twitter were around, people would make memes about it.
(Instead, they produced inscrutable late-night comedy about it:
“Comedian Red Skelton, host of the long-running Red Skelton Show, joked
about what one Miltown in his pillbox said to another. “I feel so terrible I
think I’ll take a Perry Como,” a reference to the Italian American crooner
whose soothing, velvety ballads were the butt of many jokes. “Miltown is
now coming out in four strengths,” teased NBC’s Bob Hope. “Quiet, very
quiet, rest in peace, and Perry Como.”” Tone, Age of Anxiety, 65)
It was massively popular, amounting to an unbelievable 1/3 of all prescriptions in 1957 (Tone xvi). Miltown gave way to the more known benzodiazepine class of drug, of which diazepam (Valium) is the best known.
Without getting too much into pharmacology which I don’t quite understand, each of barbiturates, Miltown, and benzos (and alcohol, seemingly; and I’m going to sloppily discuss Miltown and benzos as one and as ‘benzos’, which isn’t quite right) work on a so-called GABA receptor in the brain (or somewhere in the brain near GABA receptors). GABA receptors transmit GABA (a neurotransmitter, like dopamine and noradrenalin), and it is one of the main inhibitory neurotransmitters (by contrast, noradrenalin is one of the main excitatory ones), where that in turn means that it slows down communication between neurons (very very roughly!!), which somehow leads to feelings of relaxation.
The point is these drugs work marvellously. But in fact too marvellously. I said above that it seems as if benzos function a bit like alcohol; now imagine that your doctor prescribed alcohol (hangoverless alcohol!) that you could take any time of day or night to relax you, something reasonable owing to the discovery that so being unrelaxed is in fact pathological: a sign of illness that deserves, needs, to be treated.
People got addicted. And with addiction goes withdrawal, and the withdrawal of these GABA-ergic drugs seems to be uniformly brutal. As early as 1960 studies studies were showing that on being stopped abruptly, users experience extremely unpleasant symptoms such as anxiety, sleeplessness, nausea, and even up to hallucinations and death. (In fact, it’s widely recognised that benzos along with alcohol produce the worst withdrawal symptoms of any drug.)
These studies didn’t become public knowledge but stories began to appear in the media: people telling about their unexpected bad experiences of habituation and withdrawal.
And so, things changed. From 1975 to 1979 prescriptions about halved, from around 60 to around 30 million (Tone 205); prescribing became more proscribed (fewer repeat prescriptions without followups, etc.), and possession became a felony.
The story isn’t simple. After 9/11, per Tone, prescriptions increased, and 40 million prescriptions were written for benzos in 2006. So benzos haven’t been entirely eradicated (although scaling for population there’s certainly been a big decrease), it’s nevertheless surely true that the benzo era of the 1950s to 1970s is over and they are considerably less legally available than they were.
Benzos provide us with a first path that could happen: if in the passing of time it becomes apparent that amphetamines are riskier than seemed, and users start complaining about them, we might see tighter restrictions placed on them.
And those tighter restrictions on supply might just lead to a slackening of demand. On my reading of the history — and here’s the first place where I might be just missing a crucial piece of story — as benzos became less accessible, people didn’t have recourse to substitutes to make up for it. We don’t have stories, as far as I can tell, of the proverbial housewife cut off from valium seeking a substitute either in the legal drug market (the obvious place to look being alcohol), or, equally importantly, in the illegal market.
In my view, this feature is important. Not only was supply cut, but demand didn’t move to substitutes. But that just opens up more questions: why didn’t demand move to substitutes? In order to begin to answer that question, let’s consider our second path: criminalization.
Sabbha dukkha, the Buddha says: all kinda sucks: it’s unpleasant, unfun. Opiates are very good at making it unsuck, by dampening pain. As with stimulants, the exact story isn’t quite known. As with stimulants, opiates manipulate dopamine, but they do so in a different way, yielding a different set of psychological effects, one of which is an insensitivity to both physical and mental pain. William Burroughs talks of his years of opiate addiction as having left behind factual memories devoid of any emotional component, and there is good reason to think that the explanation is just that he was numbed to emotions. (Numbed is perhaps wrong: one is aware of pain when on opiates, they say, it just doesn’t bother one.)
The use of opium probably goes back as long as human history does. The poppy papaver somniferum has a sort of sap, ingestion which makes you high. That’s opium. Chemically treating the sap, you can potentiate the active chemicals which include morphine and codeine; doing this, you can get diacetylmorphine, aka heroin.
There are other opiates, the most notorious of which is oxycodone. Oxcodone is a bit stronger than morphine albeit weaker than heroin, and starting in the 90s was prescribed an unbelievable amount as Oxycontin.
This was because it was claimed that it managed to kill pain but by some magic didn’t lead to addiction. Notoriously, a 1980 letter in the New England Journal of Medicine suggested that addiction wasn’t a problem with opioids when used for chronic pain. Purdue pharmacy, who brought Oxycontin to market, lent on that specious peer reviewed publication to have its reps tout their drug as a solution to pain without the problem of addiction: fewer than 1% of people, they said, who used opiates correctly got addicted.
It pretty soon became apparent that oxycontin wasn’t fit for purpose (The story is very familiar, and I’m cribbing from my memory of Empire of Pain). People in genuine pain needed more of it than was claimed, and people absolutely got addicted.
Now, if the danger of stimulants is cardiovascular and perhaps psychosis in extreme cases, and the danger of tranquillisers is the withdrawal, the danger of opiates is that many of them lead to respiratory depression which can cause death.
This is especially risky when addiction is present, for several reasons. For one, addicts often try, or are forced, not to take their substance, whether because of the desire to be clean, lack of money, or imprisonment. This can cause tolerance to reset such that an amount that formerly would just have gotten them high kills.
And for another, and for some of the same reasons, the addict in withdrawal will do whatever it takes to stop it, such as using other substances. It’s thus of crucial importance that substitutes of a given addictive drug are somewhat tolerable.
So far, the story seems similar to the benzo story: life-improving drug is revealed, with a little time, to be dangerous. The big problem, and the source of the tragedy that is the opiate epidemic, is that whereas — again, on my reading — former benzo users didn’t seek substitutes, oxycontin users whose supply was cut off did. And that lead to fentanyl.
Fentanyl is a synthetic opioid, one that doesn’t derive in any respect from a plant or natural substance. It is 100 times more powerful than heroin, such that even a very small amount can kill. While it has medical use in anaesthesia, as a recreational substance it’s too risky. And yet it has merits from the point of view of drug dealers: being so potent, a small quantity is a lot of doses and a lot of profit, and moreover small quantities of something are liable to be shipped without being noticed. In addition, until a few years ago, it was legal for Chinese companies to produce it and its precursors, and, so the story goes, the main route that brought fentanyl to the US was from Chinese labs to Mexican cartels. (Apparently — and this is straight from Wikipedia — as of this year it’s being manufactured in Mexico).
Regardless, fentanyl hit the US after Purdue did. As legislators and doctors tightened up opiate prescriptions, people who hit gotten addicted got sick and looked to ease the sickness, and took to sourcing their substance illegally.
Note that this is where the story diverges from what we saw with the benzos. As I am telling it, when the benzo crackdown happened, people didn’t substitute: they didn’t look around for a benzo alternative, legal or no, to get over their withdrawals. Why that is is unknown, but it’s simply a fact that the opiate users did seek a substitute, and unfortunately, that substitute happened to be incredibly dangerous.
The opioid epidemic is truly a tragedy: humans at their worst and most unlucky. But for our somewhat more abstract purposes we need to understand why: why did the opiate users seek to ward off their unpleasant withdrawals in a way that the benzo users of the 70s didn’t? If we can answer that, we can put more flesh on hypothetical futures where amphetamine is cracked down on.
Here’s my theory of substitution: you get substitution whenever the substitute is available. It’s a very naive theory, but as far as I can tell it explains things. We didn’t get substitution for the benzos simply because there were no GABAergic drugs that the average person could access in the 70s. We did get substitution in various times and places (prohibition, various times in Russia), because moonshine or samogon can be made by most anybody.
And in the case of opiates, substitution was made available by massively increased connectivity. That connectivity brought medical research, that had discovered scores of novel pharmaceutical substances, to the hands of the many; and that connectivity enabled suitably connected people to acquire the raw materials necessary for making the substances and to send on the finished product via a globalized shipping infrastructure. The difference between the 70s and now, and our different drug outcomes, is to be explained with the internet and the shipping container.
Substitution of amphetamine?
The crucial question resulting from this is whether in the case of a crackdown people will have recourse to analogue stimulants. Such stimulants are available in the form of research chemicals like — to take one of very many examples — ethylphenidate, which as recently as a decade ago was widely available online and is analogous to Ritalin, or even wakefulness promoting drugs like Modafinil.
So substitution will be possible. Will it be actual? According to the view I’m defending, yes. We’ve opened the pandora’s box of connectivity that will enable analogues to be studied and shipped and taken.
(Here you might object: it’s ridiculous to think that the mere presence of substitutes implies their widespread use. Surely, you might say, the desirability of the drug is a factor: in particular, you might think the subjective unpleasantness of opioid withdrawal spurs users to seek substitutes while amphetamine withdrawal isn’t markedly bad and so might not spur its users to substitute.
I basically agree, but I think defending a more extreme position is helpful for viewing a topic in a new way. And note there are subtleties in the vicinity. Imagine the scenario: amphetamine becomes illegal, but isn’t substituted because its users prefer to give it up than break the law. Makes a lot of sense! But note that a main reason for illegalization in the first place would be addictiveness, and so we can reason as so: amphetamine becomes illegal, so it must be very addictive; but it isn’t substituted for, so it mustn’t be that illegal. That’s fine and inconsistent, but it does show that for the possibility considered in this paragraph to obtain we need a sort of goldilocks level of addictiveness.
Nevertheless, my considered opinion is that the subjective pull of a drug should figure more in a apt model than it does in mine.)
Before going on, we should ask the normative question as to whether these changes would be bad. Imagine use gets curbed but people don’t want to give up their stimulants and so seek a substitute in the illegal market. Would this be bad?
There’s of course some reason to be wary: exactly this move lead to the fentanyl disaster. But this seems like a case where lightning probably won’t strike twice: fentanyl is catastrophically bad in a way that other analogues probably won’t be. The switch to an illicit market for stimulants, while obviously not without risks, probably isn’t a massively worrying possibility.
What about eradication? Would that be bad? That question ultimately depends on how bad a malady ADHD is, and how much stimulant drugs change the lives of sufferers. To answer that, you need to weigh the avowals of ADHD sufferers, or try to somehow quantify productivity gains against the fact that ADHD seems culture-bound in a way that other illnesses don’t; if something is impeding a substantial percentage of Americans’ performance, and that impediment is subsequently removed, one would expect them, relative to countries where the impediment is presumably present and not removed (i.e. countries without large numbers of people medicated for ADHD), to exhibit relative performance boosts. One would expect, for example, that US college students after the Adderall revolution improved relative to their unmedicated Chinese or Indian peers (who one might presume are as susceptible to ADHD).
These are hard normative questions, and here as elsewhere the thing I’m confident about is not any particular position, but just that this captures something like the logical space of possible futures. With that said, let’s turn to the next possibility.
The two options we just considered involve cutting back on ADHD med prescriptions. Let’s consider now a couple of options that don’t involve that.
The most extreme would be that stimulants like amphetamines become generally legal. A rough analogue here would be weed (or even Viagra). Let me introduce this with a paragraph from ChatGPT.
Over the last quarter century, cannabis has undergone a dramatic transformation in its legal status in the United States. This shift began in the late 1990s when California became the first state to allow medical marijuana use, passing Proposition 215 in 1996. This historic decision marked a turning point in the perception of cannabis as a legitimate medical treatment. Subsequently, a wave of other states followed suit, enacting their own medical marijuana laws. The early 21st century witnessed a growing acceptance of the medicinal benefits of cannabis, leading to a cascade of legislative changes. This momentum culminated in 2012 when Colorado and Washington became the first states to legalize recreational cannabis, signalling a broader shift towards full legalization. Since then, many other states have embraced the idea, recognizing both the economic benefits and the need for sensible regulation in the era of changing attitudes towards cannabis. Today, a significant portion of the United States has legalized medical and recreational cannabis, reflecting the evolving landscape of marijuana policy over the past quarter century.
So let’s explore the possibility. The first thing to note is that the weed case would only be a rough analogy. For a start, the medical marijuana system was considerably different from the current use of amphetamines for ADHD. Notably, one didn’t require, in the US, a prescription, as opposed to a letter from the doctor, valid for one year, that merely suggested the patient may benefit, in general, from medical weed (no dose was indicated).
A second important disanalogy is the social justice one. The pro-legal-weed argument is, to a large extent, that the law and its implementation concerning weed is absurd, saddling people caught possessing even personal amounts to dealing with the legal system and possibly a criminal record, a fact compounded by the extreme racial inequality in the prosecution of that law.
And a third disanalogy arises from the relative strength prior to the putative legalization of the illegal market. A solid reason for legalization of weed is that it’s sold illegally a lot and so it makes sense to make it legal and tax it; but, as far as I know, the market for legal amphetamine isn’t massive, and so the tax revenues foregone aren’t going to overly trouble law makers.
In sum, the case for legalization seems dubious to me. The main pressing reasons — social justice and tax — just don’t seem to be present. If amphetamine is to remain in our society, I think, it will do so as a controlled medical substance. What would that be like?
The path of least resistance: the status quo
Finally, let’s consider the status quo position: the amphetamine business and market remains the ‘same’. I use same in quotes because it takes a bit of work to say what that amounts to. Discussing this is of inherent interest, so let’s do it.
There are various things that can be meant by ‘same’. One unlikely one is that the percentage of the population with a prescription remains the same. If that were so, then the pattern in the data would be that between 2017 and 2021 prescriptions rose by 30% but then subsequently they just stopped growing. And that would imply
But it’s not only unlikely but impossible that the rate of change continue: if the number of prescriptions increases by a third every five years, soon everyone in the US will have a prescription. This is not a comfortable stopping position: if it happens that a vast chunk of the population takes amphetamine, the case for saying it’s treating a malady is weakened.
And that yields an interesting conclusion: if amphetamine is indeed a legitimate treatment, then its use is going to have to stop growing soon enough.
I billed this as a choose your own conclusion essay. In order to do that, take a look at this tree:
Work out which points you have confidence in, based on the evidence I presented, or — better! — based on your own knowledge which I overlooked. For me, for example, I know that I think the legalization view is unlikely (because I can’t see any reason for it, and it doesn’t seem analogous to the weed case), while I think the case for cracking down is strong, simply because amphetamines seem relevantly similar to benzos and opiates. Moreover, I am quite confident in substitution, in light of the availability of alternatives as well as the connectivity that can lead to their production. I shade the bits I think likely or unlikely to add constraints:
And then I throw numbers at it based on my constraints:
These are my subjective probabilities based on what I’ve presented for the future of amphetamine, and we can use them to answer our question, repeated:
- In 5 years, will functionally amphetamine-like stimulant use in the US be lower, higher, or roughly the same as it is now?
My answer is: roughly the same. Most of my probability is on either the steady model, or the criminalization model. Hopefully it’s intuitive that on the steady model use will remain roughly the same. I think we also get it for the criminalization model: the market continues to expand for a while; then authorities step in and make it illegal; many leave now that it’s illegal, and so more or less the expansion is wiped out. Moreover, the newly illegal product doesn’t get many new comers, and so we remain at roughly a constant prevalence.
There’s much more to say. I haven’t really talked about some of the most interesting topics — the role economics plays in making drugs available; or the role of alcohol, which makes of old fruit and grains a portable and valuable mind-bender; or of coffee, which shows, in the slight but constant benefits it brings many of us, what a good drug can be. Discussing these would lead to more complicated and adequate theories, but must remain for another day.