Early classification attempts
By the mid-1800s chemists had discovered over sixty elements, but no one could see the pattern connecting them. Some elements behaved almost identically — sodium and potassium both fizz violently in water — while others seemed to share nothing at all. Three chemists, working across half a century, each tried to bring order to this growing list using the only reliable number they had: atomic mass. Johann Dobereiner spotted small families of three. John Newlands heard a musical pattern repeating every eighth element. Dmitri Mendeleev built a full table bold enough to predict elements no one had yet discovered. Each attempt captured a real piece of the truth, and each broke down at a different point — and tracing exactly where each one failed is what finally led chemists to the modern periodic table built on atomic number, not atomic mass.
Dobereiner, Newlands and Mendeleev each classified elements by increasing atomic mass, and each classification worked well for a limited range of elements before breaking down, with Mendeleev's system coming closest by leaving gaps for undiscovered elements and predicting their properties correctly.
The core idea
Dobereiner's Triads (1817-1829). Johann Wolfgang Dobereiner noticed that certain sets of three chemically similar elements, arranged by increasing atomic mass, showed a striking numerical relationship: the atomic mass of the middle element was very close to the average of the atomic masses of the outer two.
where: the "1st" and "3rd" elements are the lightest and heaviest members of the triad, and the "middle element" is the one in between, all three chosen because they already share similar chemical behaviour. For lithium, sodium and potassium (masses 7, 23, 39), the average of 7 and 39 is 23 — matching sodium exactly. Dobereiner found only a handful of such triads among the elements known at the time, so the pattern could not organise the full list.
Newlands' Law of Octaves (1866). John Alexander Reina Newlands arranged the 63 known elements in order of increasing atomic mass and observed that every eighth element resembled the first, echoing the way the eighth note of a musical scale repeats the first note an octave higher. Counting from lithium, the eighth element is sodium — and both are soft, highly reactive, monovalent metals that form similar chlorides. This repeating-by-eight pattern held reasonably well for the lighter elements but collapsed once heavier elements such as iron entered the sequence.
Mendeleev's Periodic Law (1869). Dmitri Mendeleev arranged all known elements by increasing atomic mass and proposed a general law: the physical and chemical properties of elements are a periodic function of their atomic masses. What set Mendeleev apart was his willingness to leave deliberate gaps in his table wherever the known elements did not fit the pattern, confident that undiscovered elements would eventually fill them. He even predicted the approximate atomic mass and properties of three of these missing elements — eka-silicon, eka-aluminium and eka-boron.
Limitations to remember: Dobereiner's system covered too few elements. Newlands' octaves broke down beyond calcium and he sometimes forced two elements into one slot. Mendeleev's mass-based ordering had no fixed place for isotopes, and it occasionally placed a heavier element before a lighter one to keep properties in the right group — cobalt (about 58.9) had to sit before nickel (about 58.7). That single anomaly was resolved only later, when Henry Moseley showed atomic number, not atomic mass, is the true basis of periodicity.
The goal is to verify, using only arithmetic, whether three chemically similar elements form a genuine Dobereiner triad by testing whether the atomic mass of the middle element equals the average of the atomic masses of the outer two elements to within about one unit, since this numerical test is exactly what nineteenth-century chemists used before atomic structure was understood at all.
Step 1. Identify three elements that already share obvious chemical behaviour — similar reactions, similar formulas for their compounds, similar physical appearance — since Dobereiner's rule is a check on chemically related elements, not a search through random ones.
Step 2. Order them by increasing atomic mass and label the lightest as element 1, the heaviest as element 3, and the one in between as element 2.
Step 3. Compute the average of the outer two masses: .
Step 4. Compare this average with the actual atomic mass of element 2. If the two values are very close (within roughly one atomic mass unit), the triad is confirmed both chemically and numerically; a large mismatch means the grouping fails the test even if the elements look chemically similar.
Worked example
Problem: Chlorine, bromine and iodine were one of Dobereiner's original triads. Given atomic masses chlorine = 35.5, bromine = 80, and iodine = 127, check whether they satisfy Dobereiner's rule, and state one shared chemical property.
Step 1 — Order and label. By increasing atomic mass: chlorine (35.5) is element 1, bromine (80) is element 2 (the middle element), iodine (127) is element 3.
Step 2 — Compute the average of the outer two.
Step 3 — Compare with the middle element. The calculated average, 81.25, is very close to bromine's actual atomic mass of 80 — a difference of only about 1.25, well within the rough agreement Dobereiner accepted.
Step 4 — Confirm chemically. Chlorine, bromine and iodine are all non-metals of similar reactivity that form single-charge ions and combine with sodium in the same 1:1 ratio, giving , and . This shared chemistry supports the numerical grouping.
Conclusion: chlorine, bromine and iodine form a valid Dobereiner triad, since 81.25 (calculated) matches 80 (actual) closely and all three share the same chemical family behaviour.
Common traps
- Writing Dobereiner's rule as "sum" instead of "average" — the rule takes the mean of the outer two masses, it does not add them directly.
- Treating Newlands' "octave" as a literal claim about music rather than a numbered-position analogy for a repeating pattern every eighth element.
- Saying Mendeleev classified by atomic number — he used atomic mass; atomic number as the true basis came later with Moseley.
- Forgetting that Mendeleev's correctly predicted elements (germanium, gallium, scandium) are the strongest evidence examiners expect you to cite for his system's success.
Ask the tutor: if you want to practice checking more triads or trace exactly how Moseley's atomic number resolved the cobalt-nickel anomaly, just ask your AI tutor to walk through it step by step.
Why did Dobereiner's triads work for only a few elements instead of all of them
Dobereiner's numerical trick only appears in a triad if the three elements already happen to share similar chemistry and their atomic masses happen to increase in roughly equal steps. Most known elements in his time did not fall into such neat threes — many stood alone chemically, or their mass gaps did not match. So while the pattern was real for groups like lithium-sodium-potassium and chlorine-bromine-iodine, it was a local coincidence rather than a rule covering all elements, which is exactly why chemists kept searching for a broader system.
If Mendeleev used atomic mass and it caused the cobalt-nickel problem, why is he still considered the most successful
Mendeleev prioritised chemical properties over strict mass order whenever the two conflicted, which is precisely how he correctly placed cobalt before nickel despite cobalt's slightly higher mass. His table's real strength was not perfect mass ordering — it was leaving deliberate gaps and predicting exactly what the missing elements would look like. When germanium, gallium and scandium were discovered later and matched his predictions almost exactly, it proved his underlying pattern was capturing something real, even though the mass-based anomaly still needed atomic number to fully resolve.
Is Newlands' Law of Octaves basically the same idea as Dobereiner's triads
They are related but not the same idea. Dobereiner grouped elements in threes and checked a numerical average, a small local pattern. Newlands tried something bigger: arranging all 63 known elements in one continuous sequence by mass and looking for a repeat every eight positions, closer in spirit to Mendeleev's later full table. Newlands' attempt is really the bridge between Dobereiner's narrow triads and Mendeleev's complete periodic arrangement, which is why the syllabus presents the three in that historical order.