An article from 2012 on structural realism and the periodic system

25 views
Skip to first unread message

ERIC SCERRI

unread,
Jul 14, 2025, 3:21:12 AMJul 14
to PT...@googlegroups.com
118. New Scientist. Realism. .pdf

johnmarks9

unread,
Jul 14, 2025, 10:51:34 AMJul 14
to Periodic table mailing list
Dear Eric,
Good to have you back - and with a typical thought-provoking philosophical paper!

Karl Popper stated that even if we stumble upon the truth, we can never know it. All we can know are Immanuel Kant´s phenomena. We can never know his noumena, viz. reality.

We have evolved with a brain that can understand the behaviour of things bigger than atoms and smaller than stars. But quantum mechanics and relativity are outside the realm of our experiences during evolution and therefore remain inherently incomprehensible, as Richard Feynman averred.

I don´t think theories are “relative” in the sense of the anti-realists, but they are all provisional. The force of the anti-realists´ argument is the realists` predilection for belief, one of the chief enemies of The Enlightenment. “I prefer questions I cannot answer to answers I cannot question” is Feynman´s appropriate retort to the realists.

Currently, classical physics is seen as an approximation and the sharper resolution of relativity and quantum mechanics simply shows that. But since relativity and quantum mechanics are outside our evolved understanding of experience, they remain controversial but only in a philosophical sense. Structural realism may claim, like Galileo, that the book of Nature is written in mathematics, but mathematics itself is a constant source of invention, from Abraham Robinson´s infinitesimals to the latest dimension of "string theory". They seem to be just rationalizations – and you can take your pick from which theory appeals to you most, in the absence of empirical tests. Group Theory now seems to be the fashion exploring the periodic table. Extending this particular theory of structural realism to biology seems unnecessary, for the distinction of vis viva turns out to be readily accounted for simply by using Bayesian statistics instead of Maxwell-Boltzmann statistics.

Regards,

John


On Monday, July 14, 2025 at 9:21:12 AM UTC+2 scerri wrote:

Jess Tauber

unread,
Jul 14, 2025, 6:56:45 PMJul 14
to johnmarks9, Periodic table mailing list
Hi, folks. Was there supposed to be a link to the 2012 piece?

Jess Tauber

--
You received this message because you are subscribed to the Google Groups "Periodic table mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PT-L+uns...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/PT-L/327a1cc0-5e05-4401-a047-ed5b45049c6bn%40googlegroups.com.

Jess Tauber

unread,
Jul 14, 2025, 6:59:01 PMJul 14
to johnmarks9, Periodic table mailing list
According to Google's AI, Eric was writing about this issue in 2021, rather than 2012.

Jess Tauber

Jess Tauber

unread,
Jul 14, 2025, 7:36:34 PMJul 14
to johnmarks9, Periodic table mailing list
Based on what I' reading now online, basically my entire approach to the mathematical relationships of particle numbers and energies in quantum shells (not to mention my linguistic findings) conveniently seem to fall under the rubric of 'structural realism', which I don't remember my professor ever bringing up in his course on philosophy in science (this was back in the early 1980's, though obviously it has been on the minds of those concerned with the field for very much longer than that). I'm also thinking about Max Tegmark's claims that all of material reality is just mathematics (not sure how well THAT'S doing these days, however).  It's one thing to have descriptions of the interactions between the basic parts of some construct, another to have an understanding of where those parts come from in the first place, and how 'real' they are beyond the abstract.

IIRC theoretical work on the rise of number systems started with particular cases (such as making little clay figures representing the cattle I was trading back in barter-economy times, one figure for each real cow) for record keeping (as receipts), themselves kept inside hollow clay balls created for the purpose. Not sure how universal this was, since all culture have counting numbers to some degree or other (the Yahgans, whose language I studied for many years), had only exact numeral designations for one, two, and three. The word for 'no/none' served as 'zero'. All other numerical designations were circumlocutions based on body parts (for example 20 being the sum of all the toes and fingers), or were inexact approximations (for instance from 10 to 20). I'm pretty sure no receipts for trading were created during the time when the language (whose last fluent native speaker, who was in her 90's, passed away a couple of years back) was in its heyday.

Yet even people with no extensive formal means of tracking numbers can observe the natural world and adapt their lifeways to survive in their environments. They understand the notions of relative degrees of heat, or of personal energy, light levels, weights, and so on. Most peoples without modern technology had well-developed systems (in their own languages) for keeping track of their geographical or temporal positions relative to events or other persons or things (often much more involved than the systems found in the languages of 'modern' industrialized cultures- and terms in these systems were often OBLIGATORY in conversations, just as were so-called 'evidentials', terms that reported HOW you knew or trusted that something reported was true or not, based on your having been an eyewitness, or having heard it yourself, or did your neighbor just tell you in passing, did you just suspect or hope it to be true, etc. Yahgan had a great number of such forms. What it LACKED, however, were many of the relatively generic terms we take for granted in our own languages. Yahgan terminology was nuts about SPECIFICS- in actions, states, named things, etc. No generic terms for all plants or animals (but some broad classes were recognized, such as deep-sea fishes versus those found in-shore, or big heavy birds that could feed a family versus smaller one, but there were no terms for generic 'fish' or 'birds'. Seasons were named after significant events, such as spawning upriver of certain types of fishes from the sea, or the changing of the weather (though the boundaries between these could be rather fuzzy). They did recognize when the sun changed direction in its journey longitudinally in the sky as being significant.

Jess Tauber

johnmarks9

unread,
Jul 15, 2025, 4:44:11 AMJul 15
to Periodic table mailing list
Jess,
The description you post on Yahgan is a contingency of their life-style and environmental demands. Aslian (a Mon-Khmer language spoken in the interior of the Malay peninsula) has a myriad words for every different kind of smell, because sight is not so useful if you can seldom see beyond the dense foliage surrounding you. Nor is hearing so helpful with the constant rustle of vegetation. Incidentally, Yahgan has a related neighbouring language, which they call Alacaluf, spoken in Chilean Patagonia. The speakers themselves call their language Kawesqar and number about 5,000, centred around Jetarktétqal (Puerto Eden). So you can still go and speak it, Jess. :)

Jess Tauber

unread,
Jul 15, 2025, 9:27:43 AMJul 15
to johnmarks9, Periodic table mailing list
Actually, Kawesqar (various spellings are accepted) is NOT related to Yahgan, except areally. They share some borrowings in both directions. There was a big glacier lying between their territories that you had to paddle around, so interactions were relatively rare, and brief. My good friend, the late Oscar Aguilera (who passed last year from cancer), was the world's authority on the linguistics of Kawesqar. Yahgan is (was) an isolate, meaning it had no close genetic relatives in the region. Kawesqar, on the other hand, was related to a number of other languages forming a family. So far as I had been able to determine over many years of study, Yahgan's nearest (more obvious) relations were the Salishan languages from the Pacific Northwest region of North America, covering an area roughly the size of France. This relation would be ancient, recoverable mostly by comparison of Yahgan vocabulary and grammar with reconstructions of Proto-Salishan, which, luckily for me, already exist. As for Yahgan speakers- there are no more fluent ones (except perhaps myself, haltingly), and Kawesqar is in the same boat (no pun intended), and Oscar was the last speaker there. One of his last Kawesqar informants lived for many years in New York City (undoubtedly in today's political climate he'd likely have already been grabbed by masked ICE agents and deported to Timbuktu).

By the way, the description I wrote earlier about Yahgan having few generic terms could be duplicated for quite a few hunter-gatherer groups that still exist. Interestingly, their lack of emphasis on numerical specification seems to go in the opposite direction. I wonder why that might be so. Yahgan is very rich in so-called 'serial verbs'. These are compound verbs that are chained together (usually in temporal sequence) that are then derived and inflected as all one unit. For instance 'look-see-hit-die'-past tense, would be teki-alagu:-aku:-pvn-ude: (where colon represents vowel tenseness, and teki= 'look', alagvna (v-schwa, combining form replaces -vna with -u:), aku: 'strike', and pvna 'be dead', the latter dropping the -a when adding past tense marker -ude:  So essentially 'I looked for and saw (someone, something alive), then hit and killed (him/it). Some of these serialized formations could be very elaborate (and twice as long in terms of numbers of free lexical roots involved). So far as I know the other languages in the region don't operate in this fashion at all, and longer stem forms that might have been originally compounded or derived are now nearly completely opaque to historical analysis, though there are hints here and there. Salish languages, by the way, also don't generally serialize in this manner, though they have elaborate stems compounded with so-called 'lexical suffixes' which cover a variety of semantic domains- primarily things like body parts, or commonly constructed items used for everyday tasks. This type of inclusion is called 'incorporation' by linguists, and is relatively common in North America (though rather rare in Meso- and South America).

Now, a quick note on Aslian- it is part of the Austroasiatic language family (I'm currently working on another member of that family, called Bahnar, spoken by 'Montagnards' (hill tribes in Vietnam) who helped fight with the US against the Communist North during that conflict back in the 1960's and '70's). The dictionary of that language (covering forms from a variety of dialects) has thousands of 'expressive' terms that cover very specifically notions of posture, gait, shape, size, texture, distributions (in time and space), moods and other mental states, as well as smells, tastes, and other sensations (as for instance various kinds of pain). According to the linguist Roger Blench (who has studied odor and taste expressives (a.k.a. ideophones) from languages around the world), elaborations of gustatory sensation vocabulary is relatively rare. Similarly for other 'close-up-and-personal' sensations like pain, pleasure, etc. It's the sensations whose sources are distant (like sight and sound) that tend to be most elaborated. Perhaps this relates why Pygmy languages (which are generally borrowed from their normal-height neighbors) lack very many of them, despite the fact that in general many African native languages are filled to the gills with them, especially those in open regions (even if such spaces are artificially created, such as clearings for villages or farming).

Jess Tauber

johnmarks9

unread,
Jul 15, 2025, 10:54:10 AMJul 15
to Periodic table mailing list
Thanx for the correction Jess.
Is Yamana also a Yahgan language or is it one of their Alalacaluf neighbours?
I think the Montagnards were mostly Meo (Hmong-Mien) which is why they made up most of the boat people fleeing to Hainan, etc., after the Vietnamese victory.
Interesting what you say about ideophones. I don´t know if you´ve ever tried a human gyroscope (Aerotrim®) such as are used for training astronauts, but this would need a large variety of vestibular ideophones!
Having wandered rather far from PT matters, maybe there are ideophones related to it. For example, I´m in my golden year (Z=79) . . .
Regards,
John

Jess Tauber

unread,
Jul 15, 2025, 11:22:02 AMJul 15
to johnmarks9, Periodic table mailing list
The names utilized for the Yahgan people are a mixed bag. They had no specific name for themselves (like many others, they referred to their own group as 'humans', which in their language was yamana (which also means 'canny, intelligent, healthy, able'. Yahgan is an artificial term created by the Missionary (Thomas Bridges) who worked with them the longest (from the 1850's-til 1898, when he passed). This latter term is derived from Yahgashaga (or 'Yahga Strait'), which he considered to be the center of their geographical territory (where there were five different dialects). Some less than adequately prepared supposed researchers (I use the term generously) got fooled by the nomenclatural confusion, and decided, completely unfoundedly, that Yahgan and Yamana were different entire languages (though of the same 'Yahganan' family, without any real evidence (I hate blowhards), who have tended to double down on their error no matter how gently it is pointed out to them.

There were only a few language groups as you traveled down through southern Chile (on the Pacific coast). Many are familiar with Mapuche people (Mapudungan family), where Mapu means 'earth, land'. I've done work from dictionaries on that family as well, looking for phonosemantic mappings. Then there was a group that has since died out, called Chono. Like the Kawesqar and Yahgan they were 'boat people', sleeping on land, but getting most of their sustenance from the sea (except for vitamin-rich fruits and other vegetation from the land, as well as many species of tree fungus. Chono (it is believed- there are few usable linguistic records), Kaweqar, and Ona (a.k.a. Selk'nam (the ' after the k represents, IIRC, a palatalization, so ky) form the family I mentioned earlier. The Haush, also from the same family primarily, to the south of the Ona (who were landlubbers, on the Argentinian Atlantic side of Tierra del Fuego) had a large percentage of their language filled with Yahgan borrowings. They died out long ago from European diseases they had no resistance to- the last from being forcibly confined by Catholic missionaries in cold, unsanitary conditions at the turn of the 20th century without adequate nutrition or clothes with the connivance of the government, who wanted the land for gold extraction (which turned out to be fruitless). The Haush were also mostly land-based.

Jess Tauber

Eric Scerri

unread,
Jul 17, 2025, 4:34:53 AMJul 17
to Jess Tauber, johnmarks9, Periodic table mailing list
With citations to several members of this forum including Julio Gutierrez.
 
AI and the PT study.pdf
Screenshot 2023-08-28 at 9.24.32 AM.png

Jess Tauber

unread,
Jul 17, 2025, 8:49:13 AMJul 17
to Eric Scerri, johnmarks9, Periodic table mailing list
Physicist and science-fiction author David Brin (who is a FB friend) suggested to me a couple of years back that LLM systems might be able to deliver the same kinds of phonosemantic (form-meaning) analyses I do with imitative words in human languages. While the potential is there for some level of 'wordy' (that is, lexical-type) analysis, I'm not sure their 'understanding' (and I'm using the term guardedly here) of physics and mechanics of eventive phenomena and real-world object properties is adequate to the task. Could such a system tell you how much order or disorder is involved at different hierarchical levels of a complex dynamic system or how it evolves spatiotemporally? Or do human agents have to fill in the blanks for the system? And the 'hallucinations' they are prone to having is pretty concerning as well.

Jess Tauber

On Thu, Jul 17, 2025 at 4:34 AM Eric Scerri <ericsc...@gmail.com> wrote:
With citations to several members of this forum including Julio Gutierrez.
 

Regards
ERIC 


Eric Scerri PhD



Julio gutierrez samanez

unread,
Jul 19, 2025, 9:50:11 AMJul 19
to Jess Tauber, Eric Scerri, johnmarks9, Periodic table mailing list

Thanks, Eric.

It's always satisfying to see my article published in the 2020 FOCH, included among the citations or references of subsequent works. There's no shortage of kindness from other researchers who deign to take my modest work into account. Among the references to Sebastián Consuegra - Jiménez work (2025), there's one by Katriel (2012) that I'd like to read, if you could get hold of it. This is that reference:

Katriel, J. (2012). The division of atomic orbitals with a common principal quantum number revisited: Np vs. ns. Journal of Chemical Physics, 136(14).

A big hug and greetings to all my colleagues.

Julio Antonio Gutiérrez Samanez






--
You received this message because you are subscribed to the Google Groups "Periodic table mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PT-L+uns...@googlegroups.com.

Rene

unread,
Jul 23, 2025, 2:27:52 AMJul 23
to Periodic table mailing list
Dear Eric and Julio

Unfortunately, the new article is a schemozzle.

It lists 29 numbered references, yet only 9 are actually cited in the main text. For example, Julio’s paper and the one where I’m listed as a co-author are included in the reference list but never cited in the article itself.

While the abstract claims that clustering divides the PT into four regions broadly aligning with classical divisions i.e. alkali/alkaline metals; transition metals; lanthanides/actinides; and nonmetals, the four clusters shown in Figure 7 actually comprise:
  • Cluster 0: 42% transition metals, 27% alkali/alkaline metals, 13% p-block metals, 13% metalloids, 3% nonmetals;
  • Cluster 1: 60% transition metals, 40% lanthanides/actinides;
  • Cluster 2: 37% lanthanides, 37% “heavy metals” (Sb, Te, Tl, Pb, Bi), 25% alkali/alkaline metals; and
  • Cluster 3: 100% nonmetals.
Thus, whereas cluster:
  • 0 is supposed to represent the alkali/alkaline metals, it actually has a greater proportion of transition metals;
  • 1 is supposed to represent the transition metals, 40% of it encompasses lanthanides and actinides; and
  • 2 is supposed to represent the lanthanides and actinides, it has as many "heavy metals", and zero actinides.
Elsewhere, the article variously mentions relying on either 4, 5, 6, or 8 properties, making it unclear which were actually used for clustering purposes.

The visualisations raise further questions:
  • The portrait-style dendrogram in Figure 5 is simply duplicated as a circular diagram in Figure 8a, without explanation.
  • Figure 7 bizarrely groups noble metals like Au and Pt with reactive actinides such as U and Th.
  • Chlorine and argon appear in the same cluster. Really?
  • Figure 8a assigns Sb and Te to the nonmetal cluster, but not B, Si, Ge or As. Really?
  • The caption for Figure 8a includes a second sentence that has nothing to do with the dendrogram.
  • The color-coded periodic table in Figure 8b bears no resemblance to Figure 8a.
Overall, the article is a mess and quite a disappointment. Where were the referees? How did this one get past the journal editor?

René


On 19 Jul 2025, at 23:49, Julio gutierrez samanez <kut...@gmail.com> wrote:

Thanks, Eric.

It's always satisfying to see my article published in the 2020 FOCH, included among the citations or references of subsequent works. There's no shortage of kindness from other researchers who deign to take my modest work into account. Among the references to Sebastián Consuegra - Jiménez work (2025), there's one by Katriel (2012) that I'd like to read, if you could get hold of it. This is that reference:

Katriel, J. (2012). The division of atomic orbitals with a common principal quantum number revisited: Np vs. ns. Journal of Chemical Physics, 136(14).

A big hug and greetings to all my colleagues.

Julio Antonio Gutiérrez Samanez


Jess Tauber

unread,
Jul 23, 2025, 9:10:07 AMJul 23
to Rene, Periodic table mailing list
It sounds like the young college senior biology major I knew years ago (who had already been accepted into medical school) who assured me that DNA is a PROTEIN, doubling down when I tried to talk him out of that assertion. And we wonder why medical malpractice insurance premiums are so high.....  Seems like AI (at least current iterations) won't be taking over the world anytime soon....

Jess Tauber

--
You received this message because you are subscribed to the Google Groups "Periodic table mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PT-L+uns...@googlegroups.com.

Rene

unread,
Jul 24, 2025, 4:16:56 AMJul 24
to Periodic table mailing list
On 23 Jul 2025, at 23:09, Jess Tauber <tetrahed...@gmail.com> wrote:

It sounds like the young college senior biology major I knew years ago (who had already been accepted into medical school) who assured me that DNA is a PROTEIN, doubling down when I tried to talk him out of that assertion. And we wonder why medical malpractice insurance premiums are so high.....  Seems like AI (at least current iterations) won't be taking over the world anytime soon….

I find AI can be helpful in addressing PT-related questions provided one already has some grounding in the subject.

What initially caught my eye about the new article was its claim that clustering yields four distinct “groups” broadly aligned with classical periodic divisions.

As far as I know, the classical division of the PT is into s-, f-, d-, and p-blocks. From there, one might overlay the metal-nonmetal distinction (with metalloids typically leaning toward nonmetallic chemistry).

Drilling further down:
  • Among nonmetals, we can discern four types: noble gases, halogens, metalloids, and the remaining nonmetals (H, C, N, O, P, S, and perhaps Se).
  • Among metals, we find s-, f-, d-, and p-types.
So, there’s an interesting 4-fold pattern among both metals and nonmetals. I mention this because of your interest in tetrahedrality.

There are even correspondences worth noting: noble metals paralleling noble gases; highly reactive s- and f-metals mirroring the reactivity of halogens; metalloids matching p-block metals across the zig-zag line. That just leaves the 
transition metals, which bridge the s/f metals and the p-block metals, pairing with the “leftover” nonmetals, which bridge the metalloids and the halogens. Four sets of correspondence in other words.

Coincidentally, Professor Emeritus John Wood of Goldsmiths College, University of London, has explored the prevalence of four-fold reasoning here:

Wood’s ideas tend toward speculative or philosophical interpretations rather than rigorous scientific frameworks, so they’re better regarded or read as conceptual inspiration rather than empirical authority.

Still, I presume there ought to be plenty of room for philosophical interpretations in the periodic table 🙏

René

Jess Tauber

unread,
Jul 24, 2025, 8:32:16 AMJul 24
to Rene, Periodic table mailing list
Well, remember that in the LST, with n up to 120, there are 2 f-orbitals, 4 d-orbitals, 6 p-orbitals, and 8 s-orbitals, or, 2x(1, 2, 3, 4) in the y-direction as conventionally depicted. Then the orbital lengths are 14 for f, 10 for d, 6 for p, and 2 for s, that is, 2x(7, 5, 3, 1). You remember Valery's 'perimeter rule'? The sum of the half-width of the orbital plus the height (numbers of occurrences of said orbital) is always NINE (at least in the LST we know). This directly comes from the tetrahedrality. Similarly, paired same-length LST periods contain square numbers of elements (2+2=4; 8+8=16; 18+18=36; 32+32=64). And then running sums of these squares yield tetrahedral numbers: 4; 4+16=20; 4+16+36=56; 4+16+36+64=120. You can't escape it.

Now if there is some sort of subdivision within the orbital blocks, perhaps the motif is like a 3-dimensional version of the Yin-Yang mandala? That is, each major division has its subdivisions, thus utilizing both multiplication of orbital numbers as well as division of individual orbitals. I've speculated about interrelationships of similar nature between the electronic and nuclear shell systems- In the former all LST period lengths occur twice, whereas they occur but once in the nuclear LS system. OTOH, the LENGTHS of the nuclear shells are TWICE the length of the electronic here. Do you see where I'm going here? Complementary distribution of constructional rules, based on simple mathematical progressions involving addition (versus subtraction), then multiplication versus division. Could there possibly be effects using power versus root, or even higher combinatorics (if there are any)? Would this sort of complementarity also extend to the nuclear system?

Jess Tauber

--
You received this message because you are subscribed to the Google Groups "Periodic table mailing list" group.
To unsubscribe from this group and stop receiving emails from it, send an email to PT-L+uns...@googlegroups.com.

ERIC SCERRI

unread,
Jul 24, 2025, 11:16:58 AMJul 24
to Jess Tauber, Rene, Periodic table mailing list
Please give a reference or diagram of the nuclear LST. 

ERIC

On Jul 24, 2025, at 5:32 AM, Jess Tauber <tetrahed...@gmail.com> wrote:



Jess Tauber

unread,
Jul 24, 2025, 1:32:29 PMJul 24
to ERIC SCERRI, Rene, Periodic table mailing list
1s
1p

1d2s
1f2p

1g2d3s
1h2f3p

1h2g3d4s
1i2h3f4p

Note the progression- 2 shells each consisting of 1 'orbital', then 2 of 2 'orbitals', 2 of 3 'orbitals' and 2 of 4 'orbitals'. Alternation of parity from positive to negative for each pair, parallel to the change from even to odd quantum number l (I consier 0 to be 'even' in keeping with the natural number sequence, for the sake of consistency). Numbering differs from the electronic system, obviously. Numbers of nucleons within the shells are, successively, 2, 6, 12, 20, 30, 42, 56, 72...., which are all doubled triangular numbers. 'Magics' terminating shells are 2, 8, 20, 40, 70, 112, 168, 240... which are all doubled tetrahedral numbers, of which 8, 40, 112, and 240 are twice the electronic s2 4, 20, 56, 120- every OTHER s2 electronic LS period terminal atomic number. Long ago I had also noted that the intermediate tetrahedral numbers differ monotonically from the actual s2 LS period terminal atomic numbers.  SO intermediate 2, 12, 38, 88 are, successively, , 1+1, 10+2, 35+3, 84+4 (tetrahedral numbers plus increment).

The above list is the standard sequence when only dealing with simple harmonic oscillator models of the atomic nucleus. When the spin-orbit coupling is included, then you have to take into account the so-called 'intruder' levels which drop down from the next higher shell into the lower- this is what raises the 'magic' numbers to higher values based on the nucleon counts of the intruders. The first intruder is 1g9/2, with 10 nucleons (count is always one more than the value of the numerator of the spin value, so 9+1=10). 1 g9/2 drops down 2 moves into the previous shell, 1f2p. Interestingly the drops fit right in with the structures of the previous shells- they never compromise the split-orbital  conservation of nucleon counts. That is, the p orbital, with 6 nucleons, is split by spin-orbit coupling into p3/2 (with 4 nucleons) and p1/2 (with 2). 1g9/2 jumps down over the 2p1/2 pair of nucleons (its 2-move drop) and comes between it and 2p3/2. I have only listed known FULL LS shells above. An intruder 1j15/2 is known, but no currently known members of the rest of shell its source have been synthesized. Unless heavier nuclei can be stabilized against decay or fission that may not be possible.

The sequence of drops is the successive set of doubled triangular numbers: 2 for 1g9/2, 6 for 1h11/2, 12 for 1i13/2, and 20 for 1j15/2. While nuclear physics doesn't recognize any splits or intruders below 1g9/2, the numbers are ambivalent. If one predicts a drop of ZERO moves for 1f7/2, 1d5/2, 1p3/2, or even 1s1/2, the lack of any actual INTRUSION would make it impossible to pick up evidence from the usual sorts of experiments that are done to determine placement. There might only be a slight shift of energies for the 'intruders'- which according to my readings of the Nilsson charts actually DOES take place, but is simply not registering in the minds of experimentalists. Generally intruder drops that ARE actually intrusive into the previous shell occur after THREE orbital (spin-split) partials have filled. Everything after 1d2s has more than three orbital partials, but because an s orbital splits such that you get s1/2 (2 nucleons) and s-1/2 (0 nucleons), you couldn't tell if a hypothetical 1f5/2 DID drop below 2s-1/2. There's nothing there to mark such an 'intrusion'.

Jess Tauber

ERIC SCERRI

unread,
Jul 24, 2025, 5:23:13 PMJul 24
to Jess Tauber, pt...@googlegroups.com
Thanks Jess,
I think there may be small errors in your values of 168 and 240, but I may be wrong.
Here is what I find for the shells you listed.


1s.  2
1p.  + 6     =    8

1d2s  10 + 2  =  20  
1f2p    + 14 + 6  =  40

1g2d3s   + 18 + 10 + 2  =  70
1h2f3p     + 22 + 14 + 6  =   112

1h2g3d4s  +  22 + 18 + 10 + 2 =  164
1i2h3f4p    +  26 + 22 + 14 + 6 =  232


Also wondering how the magic numbers connect with the values shown in the diagram from my book on the periodic table in which I have a chapter on nuclear structure.  ( 2, 8, 20, 28, 50 etc.) 

Screenshot 2025-07-24 at 11.22.05 AM.png

Reproduced from my, The Periodic Table, Its Story and Its Significance, OUP, 2007, 2nd ed, 2020. 

Regards
Eric Scerri

Jess Tauber

unread,
Jul 24, 2025, 8:36:38 PMJul 24
to ERIC SCERRI, Periodic table mailing list
Eric- in your counts you have 1h listed TWICE for successive shells.  1h is part of a negative parity shell, but you also have it heading the next, positive parity shell.  As you know, orbital nucleon counts shift upwards by increments of four, so s=2, p=6, d=10, f=14, g=18, h=22, I=26, and j=30. That's why your sums are off by 4 for 164 versus 168. 

The second shell containing 1h SHOULD have 1i, with 26 nucleons. That raises your 164 to 168. The next shell after should have 1j, with 30 rather than 26 nucleons, thus the shell should have 72 nucleons plus the 168 of the previous shell gives 240 rather than 232.

As for the nucleon counts in the diagram giving magic numbers- we have no spin-orbit effects listed until you get the 8 nucleons of 1f7/2 adding to the 20 count from previous shell filling, giving 28. But note that the diagram doesn't actually show any PENETRATION of the previous 1d2s structure by 1f nucleons. The remaining 12 nucleons from 1f2p, plus the 10 from the intruder 1g9/2, give the 22 needed to give the next spin-orbit magic, which is 50. Different diagrams in different texts and papers are variable in this regard, and even the treatment of shell structure by Goeppert-Mayer show variability here. Part of the nuclear 'isomer' phenomenon that might one day lead to true 'nuclear batteries', not to mention weapons with intermediate explosive power between chemical and standard nuclear weapons. The intruders don't actually become obligatory (at least in published shells I've read about) until higher up in the system. SO you have a sequence of intrusion being impossible due to acceptor shell structure to being variably instantiated and then finally obligatory, up near spin-orbit magic 82 IIRC. You can see the intruders penetrating the previous shell structure here.

Jess Tauber
Screenshot 2025-07-24 at 11.22.05 AM.png

ERIC SCERRI

unread,
Jul 24, 2025, 8:47:47 PMJul 24
to Jess Tauber, Rene, Periodic table mailing list
Thanks
But I thought I had used your list below to arrive at those numbers?

Eric

Jess Tauber

unread,
Jul 24, 2025, 8:53:23 PMJul 24
to ERIC SCERRI, Rene, Periodic table mailing list
You are correct. So I started the chain of errors. That's what I get for working on a reply too fast with only one good eye.... Oops, sorry. So, 
1h2g3d4s
1i2h3f4p
SHOULD be 1i2g3d4s and 1j2h3f4p respectively (I thought my statements about the latter shell seemed unusual- there are in fact 8 shells, with the last one only begun (there are no nuclei containing 240 neutrons).

Jess
Reply all
Reply to author
Forward
0 new messages