Category: Uncategorized

Everything You Wanted to Know about Sex*

* in earthworms

Update – Some Thoughts

Earthworms are hermaphroditic, yet they still have to find mates.  Although each worm produces both sperm and ova, a worm cannot self-fertilize. Two worms must find each other (not an easy task), exchange sperm, then later use the partner’s sperm to fertilize its own eggs.

Today my daughters noticed two earthworms in a crack of our driveway, in flagrante delicto. I grabbed my cameras, and caught a few good images. My daughters were fascinated by what they saw, and a bit troubled that their father was taking pictures of animals engaged in sex.


Here’s what was going on. The worm secretes sperm from pores on Segment 15, and can store sperm from another worm in receptacles on Segments 9 & 10. Its own ova are stored in oviducts that open on Segment 13.

Once they meet, the two worms line up in opposite directions with their ventral surfaces touching; Segments 9 and 10 of each worm are aligned with the clitellum of its partner. Setae (little hair-like structures that serve as anchors during locomotion) are extended to penetrate the body of the mate, holding the worms together. Mucus is also secreted in abundance, helping to secure the worms, and providing a moist environment for the transfer of sperm. Sperm released from each worm is carried via an external seminal groove to the seminal receptacles of the partner; after an hour or so the two worms part ways.


Later the clitellum secretes a thick mucous “cocoon,” from which the worm wriggles out backwards. As the cocoon passes over the oviduct an egg or two are released into it; then as it passes the seminal receptacles stored sperm is released into it.

Once the worm has crawled fully out of the cocoon the ends of the cocoon seal, the outside hardens, and the egg, now fertilized, develops into a young earthworm that will emerge from the cocoon in anywhere from 3 weeks to several months (perhaps longer), depending on the moisture and temperature of the soil.

I relied on these sources for information used here – I do not know this material myself:


Worm sex in cross-view 3D:


Additional thoughts:
This train of thought has been percolating in my mind all day – I’ll try to make it coherent. Sexual reproduction, that is, the need to exchange genetic material with a partner, must be very advantageous to the earthworm. The sexual anatomy is such that self-fertilization would not be difficult. When the worms copulate and exchange sperm, the partner’s sperm is stored a few segments in front of the opening from which the worm will deposit its egg into the mucous cocoon. The worm releases its own sperm from openings a few segments behind this female opening from which the egg is released; its sperm then travel externally to the storage receptacles on its partner. How hard would it be (have been) for the worm either to release its sperm to its own storage receptacles, or instead simply to release its own sperm directly into the mucous cocoon? This would avoid the risk associated with coupling with another worm in a relatively open setting (certainly more open that the worm’s burrow) for a period that typically lasts more than an hour. Clearly the benefit gained from the exchange of genes must outweigh the risk inherent in copulation.

And one more thought: one reason given for the inability to self-fertilize is that sperm and egg mature at different times, with sperm maturing first. I don’t understand this argument, as during sexual reproduction both worms allegedly use the partner’s sperm to fertilize an egg shortly after copulation; if the sperm that one worm released is not ready for prime time, surely the sperm that it received is equally immature. Something about this doesn’t ring true. Feel free to email any comments to me at

Continued reflections on the Lytro

I’ve used the Lytro for more than a year, now.  I continue to marvel at its ability to refocus, and even more at its ability to shift perspective.  In fact I use this feature a lot to create cross-view 3D images; I find them especially appealing even if many others can’t make them work.

Cross-View 3D photo of muffins

Over the past month or so I have realized that the Lytro is becoming my camera of choice.  It has its limitations (relatively low-resolution jpegs, for example), but its features that exceed those of a regular digital camera make up for them.

    • Refocusing: this is the reason that I purchased the camera in the first place – the ability to refocus macro-shots of earthworms.  I don’t often use it for this purpose, but when I do it is fabulous.

      View of antique typewriter keys.

      Jpeg export from Lytro desktop.

    • All-in-focus images: this is not officially supported by the friendly folks at Lytro yet, but it is not hard to achieve. Once a photo is perspective-shifted (one additional step after the image has been imported to your computer) you can click and drag the mouse on the image to change the perspective.  While this is happening, the entire image will be in focus. A screen shot of the image will capture the all-in-focus view.  A normal jpeg export (achieved by right-clicking on the thumbnail while in the Lytro desktop) will create a jpeg that has the focus set to the last-selected focal point.  The all-in-focus view often appears sharper overall (to my eye anyway) than the normal jpeg export, even in the region of the image on which I last focused.

      Antique typewiter keyboard

      All-in-focus version of the same photo.

    • Perspective-shift: I use this a lot to create the cross-view 3D images. The method that I used is described on my Cross-View page.  Suffice it to say that it is not difficult; about 3 minutes per image to make the 3D version.  I fully expect Lytro to build this option into the desktop in a future release, but of course this is just a guess.

    • Low-light or challenging-light situations: Although some reviewers have disparaged the Lytro in this regard, I typically find the results quite satisfactory. Even in Everyday mode the camera typically handles stage-lighting pretty well; occasionally I select the exposure region by tapping on the screen to achieve the equivalent of a (big) spot meter.  If I switch to Manual control to adjust shutter speed I can underexpose a little bit to compensate for bright faces against a dark background.  And generally I find that the 8x optical zoom (in Creative mode; 5.5x in Everyday with the most recent firmware update) coupled with the maximum 3200 iso yields a pleasing photo when I am far from a stage.
      And in those situations when the light is really too low for a good photo with any camera, the Lytro performs as well as any other.

So after a year of use, and multiple thousands of photos taken (many viewable here), I can report that the Lytro has not disappointed; in fact I am happier with it now than at the outset.  On several recent outings I have opted to take only the Lytro rather than my trusted favorite, the Canon A570.  Could it be better?  Of course; here’s a short wish-list of camera and software features that I would like to see:

  1. Higher resolution jpegs (I fear that this will have to wait for a new model of the camera).
  2. Easier export of jpegs. This should be available as a batch feature: highlight the images that you want to export and select “go.” As it is the images must currently be exported one at a time.
  3. Color and white balance adjustment.  The camera does a real good job of getting it right under many lighting conditions.  Stage lighting can cause problems – I sometimes make adjustments to the jpegs and 3D images that I create from the Lytro originals, but I can do nothing about the color of the original.
  4. All-in-focus jpeg exports. As I indicated above, I can achieve this via a work-around, but the feature could be incorporated into the desktop software.  Ideally a sliding depth of field control would be great, allowing the user to determine exactly what regions of a photo appear in focus.
  5. Wi-fi downloading of photos. When the Lytro was first introduced there were clever people who examined the patent diagrams and concluded that it contains a wi-fi chip. When will we see the ability to download to our desktop, or directly upload to the Lytro page, using wi-fi technology?
  6. And not for Lytro, but for our friends at Facebook: the ability to tag people in Lytro images in a manner that Facebook recognizes would be great!

Thanks, Lytro, for your innovation and your superb customer support. I look forward to additional products.


Liking, Disliking, and the Difficulty of Measurement


This might be a rambling post; please bear with me.  I just left an upper-level neuroscience class, filled with bright students, in which my intent had been to get us to come to terms with the difficulty of measuring emotion.  In verbal humans this is not hard (I don’t intend to trivialize the matter, but verbal self-reports of internal subjective states are a great starting point).  But what about in a nonverbal organism?  How can one measure “feeling good about something” or simply “feeling good” without verbal self-report?

Movement toward a stimulus seems on the surface to be a reasonable starting point… until one realizes that many single-celled organisms, and many plants, exhibit phototropism (moving toward light) or negative phototropism (moving toward darkness).  I doubt that there is subjective experience accompanying these movements.  So what of the rat that moves toward Purina Rat Chow? How can one be sure that there is subjective “liking” associated with this movement?

from Prince et al, (1998) Brain Research Bulletin, 47, 349.

How hard an animal is willing to work for a stimulus (pressing a lever many times for a single tiny pellet of food, for example) or how much the animal is willing to pay for a stimulus (crossing a painful shock grid to get to a receptive sexual partner) has been used as an index of drive or motivation, and surely it tells us something about the extent to which the stimulus is “liked.”  These are accepted measures of motivation in non-human animals, and generalize readily to the human condition.  But how does motivation differ from liking? Clearly the goal-object that will satisfy a drive is something that is liked, at least while the drive is present.  Can these measures of motivation be used to measure liking independently of motivation? Are they in fact measures of motivation, or do they instead measure liking (and the extent to which liking is enhanced by motivation)?

Conditioned Place Preference (CPP) has been used to determine whether an animal likes or dislikes something.  Let a rat experience the effects of cocaine in a striped box, and the effects of a placebo injection in a gray box, and later, in the absence of any drugs, the rat will choose to spend more time in the striped box.  The interpretation is that the rat associated the particular environment with the good feeling, and therefore now likes that environment.  I can accept that interpretation when CPP works — after all the stimulus that is “liked” is not there during the test, so a simple tropism can’t explain the preference.  But what about when no preference is expressed? Did the rat not care about the cocaine? Or was the rat unable to recognize the cues (a blind rat clearly would fail this test, even if it likes cocaine)? Or did the rat like the cocaine, and recognize the cues, but lack the ability to remember? A rat that can’t learn that striped box means cocaine, or learns it but fails to retrieve the association during the test, will fail to reveal via CPP that it likes cocaine.

I’m left with a sense of angst about my discipline and its ability to do what ultimately we would like to do (i.e., determine the basis of consciousness).  If an organism is capable of moving, sensing, and remembering, I think we can probably determine how it “feels” about a stimulus, using the techniques that I described above and other techniques.  However, when we beam down to the surface of an exoplanet and encounter that large blob of protoplasm (about whose ability to move, sense, and remember we know nothing) how can we hope to determine its subjective likes and dislikes?  And to be sure, I don’t have to invoke an extraterrestrial organism: how can I determine if my earthworms “like” the garbage that I feed them, or rather if they simply move toward it and eat it?

Please email any comments to


Class Demonstrations: “…sometimes it rains.”

I subtitle this post with a line from one of my favorite movies, Bull Durham: “This is a very simple game. You throw the ball, you catch the ball, you hit the ball. Sometimes you win, sometimes you lose, sometimes it rains.”  I try various demonstrations in the classroom; they often fail miserably, but “sometimes you win.”  This Emotional Stroop Effect demo worked well.  Students had to select the color of a word, and words with negative emotional significance slowed the responses compared to neutral words.  Why this happens is open to much discussion, but I was glad to see that I could replicate this finding in the classroom.

And “sometimes it rains.” I’m not calling this one a loss, because I don’t really understand it.  I tried a demonstration of relative validity in my learning class.  This is an effect that holds across all species tested so far, except for goldfish.  In learning theory parlance, two groups receive Correlated and Uncorrelated presentations of some conditioned stimuli and an unconditioned stimulus:

Correlated:             AX+               BX-

Uncorrelated:       AX+, AX-      BX+, BX-

And the relevant question is, what do the groups learn about XX is paired with the US (+) 50% of the time for both groups, and yet typically the Uncorrelated Group is more likely than the Correlated Group to give a conditioned response to X.   The explanation for this is that we attend to the best predictor of the US, and for the Correlated Group the best predictor is A because it is always paired with the US.  For the Uncorrelated Group all three stimuli, A, B, & X) are paired with the US 50% of the time, so all are equally good predictors and should therefore be learned about equally.

As an example in Pavlov’s dog, say, A is a bell, B is a light, and X is a Pat on the back.  For Correlated Dogs, Bell+Pat is always paired with food; Light+Pat is never paired with food. These dogs should learn that Bell means food; they’ll salivate a lot to the Bell but not much to the Light or Pat.  For Uncorrelated Dogs, Bell, Light, and Pat are all paired with food half of the time, so the dog should drool equally often to all of them.

My students experienced three stimuli, all presented in the form of a photo of two musicians.  A was a Green Guitar being played by one of them, B was a Purple Beard on the other musician, and X was simply the Two Guys.  Correlated Group saw Two Guys + Green Guitar paired always with the label “Rock Stars;” and Two Guys + Purple Beard with the label “Losers.”  Uncorrelated Group saw the same two photos as well as the other possibilities: Two Guys + Green Guitar labelled “Losers” and Two Guys + Purple Beard labelled “Rock Stars.”  After viewing 20 photos (half of which contained the Green Guitar, half of which contained the Purple Beard, half of which contained the label “Rock Stars,” half of which contained the label “Losers,” and all of which contained the Two Guys, with photos set up according to the Correlated or Uncorrelated conditions) the students were shown the Two Guys (no Green Guitar or Purple Beard) and simply asked “Rock Stars” or “Losers?”  The results appear in the figure.

Did it work? I don’t know – email me at wjwilson[at] with your thoughts.  The response to X, the Two Guys, differed across the Correlated and Uncorrelated Groups, so that’s good.  I had expected a difference in the opposite direction, though.  It appears that the association between X (Two Guys) and the label “Rock Stars” was much stronger for the Correlated Group than for the Uncorrelated Group, and I naively expected the opposite.  As the stimulus Two Guys was paired with the label “Rock Stars” equally often for the two groups yet I found a difference in its meaning, I suppose this reflects relative validity.  But how can I wrap my head around the direction of the effect?  If I had plotted the proportion of people in each group who answered “Losers” the results would look more like I expected (that is, X associated more strongly with “Losers” for the Uncorrelated Group), but this does not convince me that I am seeing a normal relative validity result.  If you have experience with human relative validity studies and have any thoughts on this, please let me know.

WordPress Themes