Experimental demonstration of existence of atoms

HJ Farnsworth

Greetings,

My questions below could be categorized into a mixture of “history of chemistry” and “experimental basis for chemistry”. I’m having difficulty phrasing the questions that I have, so I’m going to start by stating them as directly as I can, and then spend the rest of this post clarifying what I am looking for.

DIRECT STATEMENT OF QUESTIONS:

What experiments could be performed to conclusively show …

1. That matter is composed of discrete particles (atoms and molecules)?
2. Whether a given form of matter is elemental, as is neon gas or gold, or not, as is O₂ gas or brass (Cu + Zn alloy)?

QUESTION 1 – CLARIFICATION:

Given a material, such as water or a piece of paper, one could divide that material into smaller segments repeatedly. An important scientific question is – could this division be repeated indefinitely, indicating that matter was continuous, or would a point eventually be reached where a fundamental building block of matter was reached, indicating that matter was composed of discrete particles. Somewhere along the historical road (I know that Dalton and Mendeleev are big contributors to the story), the scientific community widely accepted the latter possibility to be the correct one.

What was the experimental basis for this conclusion? More importantly, what is a simple but rigorous set of experiments (I explain what I mean by this phrase at the bottom of this post) that would lead to this conclusion?

To put what I am looking for another way, and a way closer to what I am after, let’s pretend that I was magically transported to a society in, say, the 1700s, and I wanted to demonstrate to the society that matter was composed of discrete particles. How could I do this?

QUESTION 2 – CLARIFICATION:

Even if I have experimentally demonstrated that matter is composed of discrete particles, I have not shown that some classes of these particles (molecules, alloys, solutions, etc.) are in fact entirely composed of combinations of a more fundamental class (atoms). In certain cases, I could demonstrate that a given substance can be divided into more fundamental substances – for instance, electrolysis of H₂O divides water into H₂ and O₂, which is solid evidence that H₂O is not an element. But of course, neither are H₂ and O₂, since they are diatomic molecules. How do I know when I have reached the “stopping point” of dividing up matter into more fundamental building blocks (excluding concepts like protons, neutrons, and electrons)?

Many elements and non-elements have been in use or at least known by societies since antiquity. Copper, gold, and iron are made up of bulk groupings of single elements. Bronze, brass, steel (alloys) and rust (molecule) are not. These are all bulk materials that appear reasonably similar to each other.

What simple but rigorous set of experiments could I use to conclusively demonstrate (once again, to a society that does not have modern science or technology at its disposal) whether the discrete components of a bulk material could be broken into more fundamental components?

ONE MORE THING:

I want to clarify what I mean when I use the phrase “simple but rigorous”. When you read scientific literature, it is densely written and relies on various other sources in scientific literature. This is not unique to modern science – if you read scientific papers written in the 1800s, for instance, the situation is usually similar, referencing recently discovered phenomena and material properties from various other sources. So, historically, you get an enormous mess of interconnected experiments, out of which coherent conclusions manage to arise.

However, if today we decided to set up experiments to re-demonstrate the principles that form the core of our modern understanding of science, it would be silly to exactly copy the historical route, with all of its messiness. Since we would know what we were looking for, a much more organized set of experiments could be set up, and they would demonstrate the same principles just as conclusively.

I would qualify the historical experiments as “chaotic/messy but rigorous” – they have to be, since there is no way people could have known the optimal experimental set-ups to eventually arrive at the correct conclusions (this is not a criticism at all, the same is true today and there is no reason that it shouldn’t be). However, the experiments that could be redone to demonstrate the same scientific facts in hindsight could be set up to be “simple but rigorous”.

In other words, I’m after a set of experiments that demonstrate the concepts that I talked about above, without the messiness that came from lack of knowledge when they were originally demonstrated, but that still do not depend on knowledge of the concepts that I’m trying to demonstrate.

Anyway, the length of this post is probably reaching the limits of people’s patience, so I’ll stop. Please let me know if my questions are unclear.

Thanks for any help that you can give.

-HJ Farnsworth

DrDu

A very simple one is the oil drop experiment, where you estimate the size of fatty acid molecules from the size of a fat sheet on water corresponding to one drop.

Constant proportions of in chemical reactions in combination with the ideal gas law is also easiest explained in terms of molecules made up of a fixed number of atoms.

Nevertheless even at the beginning of the last century some physicists were not convinced of the existence of atoms (namely Ernst Mach).

MrAnchovy

There is no such thing as an experiment that "conclusively shows" anything, the most you can hope for is one that "convincingly supports" a scientific theory.

For the first theory I would choose Brownian Motion, but for the second I think you will have a problem devising an experiment that demonstrates that something cannot be done.

mfb

Atoms and molecules are not elementary particles, so it IS possible to split them into smaller parts. You will lose their chemical properties, however.

Mass spectrometry on pure substances clearly shows that your substance consists of many particles with the same, tiny mass.
Atomic force microscopy and electron microscopes give a direct way to see individual atoms.

Mass spectrometry to check the chemical composition, chemical reactions and/or mass spectroscopy to check the number of atoms per molecule (as in O₂). The heat capacity together with some knowledge about quantum mechanics allow to see that oxygen consists of molecules, too.

Those techniques (apart from chemical reactions) were not available in 1700, however.

That is tricky. You might be able to establish the period system, and show how many elements fit nicely in that system. This will not work for metals, however. For alloys, you can show that they have more than one element in it via chemical reactions.

At the same time, it might be interesting to establish models of the atoms with the Rutherford experiment and some research about electricity.

So you look for an alternative history, where the same knowledge was gained, just in a quicker way?

HJ Farnsworth

Thanks for the replies.

Very interesting, I did not know that the existence of the atom was still controversial at the beginning of the 1900s. I think I assumed this because things like the J.J. Thomson's experiments were in the late 1800s, and seemed to presuppose the existence of the atom. However, I did a little reading after looking at your post, and you are correct.

I've been reading around, and it seems like what you and MrAnchavoy said about proportions of atoms, the ideal gas law, and Brownian motion is correct. The (very rough) flow of scientific thought, historically, seems to be: conservation of mass in chemical reactions (Lavoisier), then law of definite (then multiple) proportions (Proust, then Dalton). Later, Avogadro made contributions using the ideal gas law, and Brown's discoveries combined with Einstein's theories about them sort of put atomic theory beyond controversy in mainstream science.

HJ Farnsworth

Hi mfb, I just saw your post also. I haven't really been able to go over or think about it much yet - so hopefully I'll be able to do that sometime in the next couple of days. Anyway, thanks for responding.

Yeah, alternative history is a good phrase for it. Basically, if every experiment that had been done gave us information that we were looking for in as quick concise a way as possible, what would the experiments that gave us information regarding the existence of atoms look like?

MrAnchovy

It doesn't work like that. Experiments only make sense in the context of knowledge that is already posessed: looking at an atom in a scanning tunneling microscope is pretty convincing in the 21st century, but to Euler or Newton it would have been no more credible than pulling a rabbit out of a hat.

DrDu

Yes, and this works also in the other direction: I once read a book about Ramon Llull http://en.wikipedia.org/wiki/Ramon_Llull
He is sometimes regarded as the founder of informatics.
He was convinced to have found a logical system with the help of which he would be able to convince any muslim of the truth of cristianism, the Lullian circle.
It is quite difficult to grasp and at least for me, only seemed to produce quite trivial statements.
(On the other hand, this is exactly what a computer does in every step).
Nevertheless it influenced e.g. Leibnitz who produced maybe the first calculator.