Practical Field Guide to Igneous Rock Classification
I started carrying a thin section kit and a streak plate to the field after my third year of mapping, and honestly I wish I'd picked it up earlier. The basic distinction between intrusive and extrusive rocks comes down to one thing: cooling rate. That's it. But applying that principle when you're standing in the rain with a hammer and trying to decide whether a rock formed two kilometers underground or erupted on the surface is where things get messy. The textbook answer says intrusive rocks cool slowly and form large crystals, while extrusive rocks cool quickly and form small crystals or glass. That is directionally correct and will get you through most undergraduate labs. In practice, texture overlap is so common that crystal size alone will fail you if you are not careful about context. Start with the hand specimen and look at the overall texture before you start arguing with yourself about whether a feldspar grain is 2 millimeters or 3 millimeters. Phaneritic texture means visible crystals, which usually points to intrusive origin, but you can get phaneritic textures in extrusive rocks near flow contacts where cooling is slower. Aphanitic texture means crystals are too small to see without magnification, which usually points to extrusive, but you can also get aphanitic textures in shallow intrusive bodies like dikes where the marginal chill zone freezes the texture down fast.
Here is the part people miss: look at the contacts. Intrusive rocks show cross-cutting relationships, assimilation walls, and chilled margins on the outside of the intrusion. Extrusive rocks sit on top of sequence boundaries, show vesicles, flow banding, and sometimes rest directly on volcaniclastic debris. A rock with no visible vesicles and no flow layering and cutting across older units is very likely intrusive. A rock sitting conformably within a sedimentary sequence with vesicular top and a glassy base is almost certainly extrusive. I spent two days stuck on a basaltic sample from the Columbia River Basalts group because the top was perfectly glassy and the bottom was porphyritic with 5-millimeter olivine phenocrysts. My initial read was extrusive flow, but the phenocryst population and the lack of any vesicle zone suggested a shallow sill rather than a lava flow. I used a combination of thin section analysis, measuring the plagioclase zoning with polarized light, and checking the surrounding stratigraphy for baked contact metamorphism. The zoning showed sustained crystallization before emplacement, and the baked contact clinched it: shallow intrusive sill, not extrusive. That single sample changed how I approach porphyritic basalts in the field.
What to Look for in the Field
Use a hand lens at 10x minimum. Look for vesicles. Vesicles are spherical or elongate cavities left by gas bubbles escaping the melt. They are extremely common in extrusive rocks and rare in intrusive rocks because pressure at depth keeps gases dissolved. If you see vesicles, the rock is likely extrusive unless you have a very unusual high-pressure shallow intrusion, and even those rarely produce well-developed vesicle networks. Check for flow banding. This is a layering of different mineral compositions or crystal sizes that records the movement of the melt before it froze. Flow banding only happens in extrusive settings. I have seen students call a banded gneiss a flow-banded rhyolite once. It was a metamorphosed pluton, not a volcanic flow. The difference is that gneissic banding involves mineral segregation under pressure and heat, while flow banding involves physical layering of the melt itself. If the banding is wavy and compositional layers pinch and swell like liquid, it is flow banding. If it is straight, flat, and controlled by regional stress, it is foliation, not flow. Look at the crystalline framework. In intrusive rocks, crystals interlock in all directions without a preferred orientation because they grew freely in three dimensions over thousands to millions of years. In extrusive rocks, early-formed phenocrysts may show alignment from flow, and the groundmass is uniformly fine-grained because the entire remaining melt solidified nearly simultaneously. Porphyritic texture with large phenocrysts in a fine groundmass can occur in both settings, so do not use porphyritic texture as a standalone diagnostic. It tells you something about cooling history, not about the final emplacement depth.
Get the Full Details
When Crystal Size Alone Fails
This is the biggest trap. Some extrusive rocks have surprisingly large crystals. Volcanic bombs can retain phenocrysts larger than 1 centimeter because they spend time in the air and cool gradually enough for crystals to grow during transport. Some shallow intrusions, like lapilli dykes and sills, cool fast enough at their margins to produce aphanitic textures that look extrusive at first glance. A diabase dyke can look identical to an andesite flow in hand sample if neither has been heavily weathered. The rock type and the textural label are not the same thing. Here is what I do when the texture is ambiguous. I measure the groundmass grain size using a grain size chart against the hand lens and record it precisely. Then I look for accessory indicators: iron oxide staining around vesicles suggests extrusive origin because oxidation happens at the surface. I check for xenoliths, which are fragments of country rock incorporated into the magma. Xenoliths are more common in intrusive settings because the magma is plumbing through existing rock, but they can occur in extrusive eruptions too, so this is not definitive. I use the presence or absence of a chilled margin at the contact. Chilled margins are diagnostic of intrusive contact with cooler country rock. If I still cannot resolve it, I take a thin section. Under the microscope, I look at crystal habits. Intrusive crystals tend to be euhedral to subhedral with clean grain boundaries and uniform zoning. Extrusive phenocrysts often show resorption rims, corrosion features, and complex oscillatory zoning that records rapid changes in the magma chamber before eruption. Groundmass texture matters too: spherulitic textures and microlite arrangements indicate extrusive. Cryptocrystalline groundmass with devitrification indicates extrusive volcanic glass that has been slowly altering.
Specific Workaround for Altered Specimens
Altered rocks are the worst. My most frustrating case was a sample from the Sierra Nevada foothills that had been hydrothermally altered enough that the original texture was nearly erased. The rock was green from propylitic alteration, full of chlorite and epidote, and I could not tell whether I was looking at a volcanic tuff or a shallow intrusive porphyry. I solved it by focusing on the altered phenocrysts. The plagioclase phenocrysts showed relict compositional zoning that was visible under polarized light even though the rock was saturated with secondary minerals. That zoning pattern matched a shallow intrusive origin, not an extrusive one. I also found a thin vein of quartz that cross-cut the sample, which is consistent with hydrothermal fluid pathways in intrusive systems rather than volcanic ones. For altered specimens, do not rely on macroscopic appearance alone. Use the thin section, check for relict textures, and look at the alteration assemblage. Propylitic alteration (chlorite-epidote-albite) is common in both settings, but argillic alteration (clay minerals) tends to be more associated with shallow intrusive systems where hot fluids interact with meteoric water.
Limits of the Method
This approach breaks down in several scenarios. Volcaniclastic rocks that have been compacted and slightly cemented can look like extrusive igneous rocks in hand sample but are technically sedimentary. Metamorphosed volcanic rocks lose their primary textures entirely. Samples from highly weathered terrains lose their distinctive surface features to oxidation and clay formation. If you are working with samples that have been tumbled or machined, you lose the natural fracture surfaces and weathering rinds that provide critical clues. In those cases, geochemical analysis is the only reliable fallback. Bulk rock chemistry using XRF or ICP-MS will tell you the tectonic setting and compositional family, and combined with field context, it resolves most ambiguities. Thin section work is still necessary because chemistry alone does not distinguish intrusive from extrusive for the same composition. A rhyolite and a granite can have identical chemistry, but one erupted and the other did not. Field mapping with careful attention to stratigraphic position, contact relationships, and associated deposits remains the most efficient first step. Thin section microscopy is the confirmatory step. Geochemistry is the last resort when everything else is ambiguous. Most samples resolve at step one or two if the exposure is good. Bad exposures, faulted terrain, and deep weathering profiles are where this system fails, and that is just the reality of working in the field.
