The Rules Inheritance Follows, and the Ways It Breaks Them

An elephant only ever gives birth to another elephant. A mango seed only ever grows into a mango tree. Whatever mechanism governs inheritance, it is clearly reliable enough that species stay recognisably themselves, generation after generation. And yet offspring are never perfect copies of their parents either: siblings can look strikingly similar, or strikingly different, and either outcome sits comfortably within what counts as normal. The branch of biology that studies both halves of this puzzle at once, inheritance, the reliable passing of characters from parent to offspring, and variation, the degree to which offspring actually differ from their parents, is called genetics. People understood, in a rough practical sense, that both existed for thousands of years before anyone understood the mechanism behind either. Farmers exploited naturally occurring variation in wild plant and animal populations for millennia, selectively breeding for traits they wanted, which is exactly how ancestral wild cows became distinct, deliberately bred lines like the Sahiwal cattle still raised in Punjab today. What those farmers lacked was any real scientific account of why any of it actually worked.

Mendel's basic pea experiments already established the core mechanism: every trait is controlled by two gene copies, called alleles, one from each parent, and when the two differ, a dominant allele masks a recessive one. Formalising the vocabulary here pays off quickly. An organism's actual genetic makeup, its specific combination of alleles, is called its genotype, written with letters like TT, Tt or tt. What that genotype actually looks like from the outside, tall or short, round or wrinkled, is called its phenotype. An organism carrying two identical alleles, TT or tt, is homozygous for that gene; one carrying two different alleles, Tt, is heterozygous. Since a heterozygous Tt plant looks identical from the outside to a homozygous TT plant, both simply tall, working out which genotype a tall plant actually has requires a test cross: crossing the unknown plant with a homozygous recessive tt plant and examining the offspring's ratio, since only a heterozygous parent will produce any short offspring at all. A British geneticist named Reginald Punnett developed a simple grid, now called the Punnett square, for laying out every possible combination of parental gametes and reading off the resulting genotype and phenotype ratios directly, a tool used constantly throughout the rest of this chapter.

Mendel's peas always showed clean dominance, one allele fully masking the other, but that turns out to be only one possible outcome, not a universal law. Cross a true-breeding red-flowered snapdragon with a true-breeding white-flowered one, and the F1 offspring are neither red nor white, they are pink, a phenotype matching neither parent. Self-pollinate those pink F1 plants, and the F2 generation splits into a 1 red : 2 pink : 1 white ratio, a genotype ratio that matches an ordinary monohybrid cross exactly, but a phenotype ratio that clearly does not, since the heterozygote is now visibly distinguishable from both homozygotes instead of hiding behind one of them. This is called incomplete dominance, and it makes sense once you consider what dominance actually depends on: whether a single working copy of an allele produces enough of its protein product to fully match what two working copies would produce. When it does, you get ordinary dominance. When one copy alone is not quite enough, the heterozygote ends up somewhere in between, exactly what pink snapdragons demonstrate.

Incomplete dominance blends two alleles' effects into something in between. Co-dominance does something different again: both alleles express themselves fully and simultaneously, with neither blending into nor masking the other. Human ABO blood groups are the textbook example. A single gene, I, comes in three possible alleles: IA and IB, each producing a slightly different sugar marker on the surface of red blood cells, and i, which produces no marker at all. Since every person is diploid, each of us carries exactly two of these three possible alleles. IA and IB are each completely dominant over i, so someone with IAi genotype shows blood type A, and someone with IBi shows blood type B. But when a person inherits one IA and one IB, something different happens: both alleles express their own marker simultaneously, neither masking the other, producing blood type AB, red blood cells genuinely carrying both sugar types at once. With three alleles available but only two present in any individual, six genotypes are possible in total, IAIA, IAIB, IAi, IBIB, IBi, and ii, but they collapse into only four visible blood types: A, B, AB and O. This is also a clear case of multiple alleles, more than two versions of the same gene existing across a population, even though any single individual can only ever carry two of them at once.

Mendel published his results in 1865, but they went almost entirely unnoticed for thirty-five years, partly due to poor publicity, partly because his mathematical approach to biology struck contemporaries as bizarre, and partly because he had no physical proof that his abstract 'factors' were made of anything real at all. Three scientists independently rediscovered his work in 1900, right around the time improving microscopy let researchers watch cell division far more closely than before, revealing structures inside the nucleus that visibly doubled and divided just before every division: chromosomes. By 1902, two scientists, Walter Sutton and Theodor Boveri, had worked out exactly how chromosomes move during meiosis, and they noticed something striking: chromosomes behave exactly the way Mendel's factors were supposed to behave. Both occur in pairs. Both segregate during gamete formation so that only one member of each pair reaches any given gamete. Different pairs assort independently of each other. Sutton put these parallels together into a single unified idea, the chromosomal theory of inheritance: Mendel's abstract factors were not abstract at all, they were physically located on chromosomes, and the chromosome movements visible under a microscope were the physical mechanism producing Mendel's mathematical ratios.

Verifying the chromosomal theory experimentally fell to Thomas Hunt Morgan, who chose an almost perfect laboratory subject: Drosophila melanogaster, the common fruit fly. It grows on simple synthetic food, completes an entire generation in about two weeks, produces huge numbers of offspring from a single mating, has clearly distinguishable males and females, and carries plenty of visible hereditary variations, all useful features for tracking inheritance across many generations quickly. Morgan ran dihybrid crosses similar to Mendel's, hybridising yellow-bodied, white-eyed female flies with brown-bodied, red-eyed males, expecting independent assortment to eventually produce the familiar range of combinations. Instead, the two genes stubbornly refused to assort independently, the F2 ratio deviated dramatically from what independent assortment predicted. Morgan realised the two genes both sat on the very same chromosome, physically tied to each other, and coined the term linkage for this physical association, plus the term recombination for the rarer new combinations that did occasionally appear, generated when homologous chromosomes exchanged segments during meiosis. Genes sitting very close together on a chromosome, Morgan found, showed very little recombination, since there was little physical distance between them for an exchange to occur, while genes sitting farther apart on the same chromosome showed noticeably more. His student Alfred Sturtevant turned this observation into a genuinely useful tool, using recombination frequency between gene pairs as a direct measure of physical distance between them, producing the first genetic maps, the same basic strategy later scaled up enormously for the Human Genome Sequencing Project.

Every trait covered so far has come with two clean alternatives, tall or dwarf, round or wrinkled. Look around at real human variation and that clean split mostly disappears: human height does not come in two categories, it spans a continuous range, and the same is true of skin colour. Traits like these are typically controlled not by one gene with two alleles but by three or more genes acting together, called polygenic inheritance, with environmental factors often contributing as well. Imagine three separate genes, A, B and C, each contributing to skin darkness, with dominant capital-letter versions adding darkness and recessive lowercase versions not. A genotype carrying all six dominant alleles, AABBCC, produces the darkest possible skin colour; one carrying all six recessive alleles, aabbcc, the lightest; and every combination in between produces a correspondingly intermediate shade, since each individual allele's contribution simply adds onto the others. There is also a case running in the opposite direction: instead of many genes shaping one trait, occasionally a single gene shapes many different traits at once, called pleiotropy. Phenylketonuria in humans, caused by a single mutated gene that fails to produce a working copy of one particular enzyme, produces two seemingly unrelated effects simultaneously, mental retardation and reduced pigmentation in hair and skin, both traceable back to that same single broken enzyme's knock-on effects throughout the body.

In 1891, a researcher named Henking noticed a peculiar structure surviving through sperm formation in certain insects, present in exactly half the resulting sperm and absent from the other half. He called it simply the X body, without any idea what it actually did. Later work identified it as a chromosome, the X chromosome, and revealed that its presence or absence was directly deciding the sex of the resulting offspring, a discovery that opened up comparing sex-determination systems across very different animal groups. In many insects, including grasshoppers, the system is XO: females carry two X chromosomes, males only one, alongside the ordinary autosomes, with sex fixed entirely by whether a given sperm happens to carry that single X or not. Humans and Drosophila both use a related but distinct system, XY: females carry two X chromosomes, males carry one X plus a smaller, distinct Y chromosome, and since every human egg carries an X by default, it is entirely the father's sperm, X-bearing or Y-bearing in a fifty-fifty split, that decides a child's sex. Birds run the whole logic in reverse, a system called ZW: males carry two identical Z chromosomes while females carry one Z and one W, making the female, not the male, the sex that determines offspring sex through her eggs. Honeybees abandon chromosome pairs altogether in favour of chromosome number itself: a fertilised egg, carrying a full double set of chromosomes, develops into a female, while an unfertilised egg, carrying only a single set, develops into a male through parthenogenesis, meaning male honeybees have no father at all, and, oddly, cannot have sons of their own, only daughters, since their sperm carries a complete chromosome set rather than a half share.

Every mechanism in this chapter so far has assumed genes and chromosomes are stable, faithfully transmitted unchanged from parent to offspring. Mostly, they are, but not perfectly, and any change in the genetic material itself, altering genotype and, potentially, phenotype, is called mutation. Some mutations affect a whole chromosome, entire segments lost through deletion or gained through insertion or duplication, disturbances of exactly the kind commonly found in cancer cells. Others affect just a single base pair within a gene, called a point mutation, and sickle-cell anaemia, covered shortly, is the textbook example. Various chemical and physical agents, called mutagens, are known to increase the rate at which mutations occur, ultraviolet radiation being one of the most familiar. Mutation matters for two very different reasons at once: it is the ultimate source of every new genetic variant a population has ever had access to, the raw material this whole thread's earlier chapters kept returning to, and it is also, when it strikes a critical gene, the direct cause of a range of genetic disorders, which is exactly where this chapter turns next.

You cannot ethically run a controlled cross experiment on human beings the way Mendel did with pea plants, so tracing how a trait or disorder moves through a human family relies instead on pedigree analysis: mapping the trait's appearance across several generations of an actual family tree, using a standard set of symbols. Disorders traceable to a single altered gene, called Mendelian disorders, follow inheritance patterns pedigree analysis can reveal clearly, dominant or recessive, and, in some cases, linked specifically to a sex chromosome. Colour blindness, a sex-linked recessive disorder affecting red-green colour discrimination, sits on the X chromosome, which is exactly why it affects roughly 8 percent of men but only about 0.4 percent of women: a man carrying just one X needs only one altered copy to be affected, while a woman, carrying two X chromosomes, needs both copies altered, an outcome her own unaffected but carrier mother makes far less likely. Haemophilia follows the identical X-linked recessive logic, disrupting a protein in the blood-clotting cascade so that even a minor cut can bleed uncontrollably, and its most famous documented pedigree belongs to Queen Victoria, a carrier whose descendants across several European royal families inherited the condition. Sickle-cell anaemia works differently, an autosomal recessive condition affecting a single pair of alleles, HbA and HbS, where only the homozygous HbSHbS genotype actually shows the disease, caused by one single amino acid substitution, glutamic acid swapped for valine, at one specific position in the haemoglobin protein's beta chain, which makes the affected haemoglobin molecule clump together under low oxygen and distort red blood cells from their normal biconcave disc shape into a rigid, elongated sickle. Thalassemia, also autosomal recessive, produces a related but chemically distinct problem, not a faulty haemoglobin molecule but simply too little of one, since mutations reduce the production rate of one of haemoglobin's globin chains rather than changing its shape.

Mendelian disorders trace back to a single altered gene. Chromosomal disorders are a different category entirely, caused by an abnormal number of entire chromosomes rather than any single gene defect. When chromatids fail to separate properly during cell division, an error called aneuploidy, an individual can end up with one extra copy of a chromosome, called trisomy, or one missing copy, called monosomy. A separate failure, cytokinesis not completing after nuclear division, can instead duplicate an organism's entire chromosome set, called polyploidy, a condition common in plants but rare and usually not viable in humans. Down syndrome is caused by trisomy of chromosome 21, an extra full copy, producing a short stature, small rounded head, furrowed tongue, distinctive palm crease and delayed physical, psychomotor and mental development. Klinefelter syndrome arises from an extra X chromosome in a male, a karyotype of 47, XXY, producing overall masculine development alongside some feminine characteristics such as breast development, with affected individuals typically infertile. Turner syndrome is the reverse kind of error, the complete loss of one X chromosome in a female, karyotype 45, X0, producing infertility from underdeveloped ovaries along with an absence of other secondary sexual characteristics. All three conditions, along with many others, can be identified directly by karyotyping, laying out and counting an individual's chromosomes under a microscope, exactly the same technique this thread's earlier chapters have already relied on to explain ordinary, healthy chromosome pairing.

Step back across this entire chapter and a clear pattern emerges. Mendel's basic rule, two gene copies, one dominant, is solid enough to explain why an elephant reliably gives birth to an elephant, but real inheritance immediately complicates that rule in every direction: alleles that blend instead of dominating, alleles that both express at once, genes that refuse to assort independently because they are physically tied to the same chromosome, single traits built from many genes at once and single genes affecting many traits at once, entirely different chromosomal logics deciding sex from one species to the next, and, running beneath everything, mutation quietly introducing fresh variation, and occasionally disorder, into every one of these systems. None of it is really a contradiction of Mendel's original insight, it is what happens when that insight, discovered from a comparatively simple pea plant, gets tested against the full, genuinely messy range of life on Earth. But everything covered in this chapter has treated a gene as a kind of black box: something that gets inherited, that dominates or doesn't, that occasionally mutates, without ever actually opening that box to see what a gene does, mechanically, inside a living cell, to turn its information into an actual protein. That is exactly where this thread goes next.