I. Memory Systems
1. Ebbinghaus (1885). First experimental research on memory. Ebbinghaus discovered the forgetting curve, and introduced "unconscious" memory as illustrated by his savings method. He introduced nonsense syllables as a research tool, and empirical evidence for some structure in memory. also set the pattern for a research tradition that to some extent extends to the present day.

2. Different memory systems? Short Term Sensory Store (STSS) or sensory register, Short term memory (STM), Working memory, long term or permanent memory (LTM or LTS). Also there may be visual memories, spatial memories, semantic memories, episodic memories, motor memories, music memories, etc. Anderson uses the data that seems to show that most measures of forgetting and improved memory seems to function by a power law as evidence that there is only one memory. I don't know whether those data are secure or not, but even if it turns out that memories all get stronger and weaker by a power law under some conditions it does not convince me that there is only one memory system. Whether there are different memory systems is currently not agreed upon by Cognitive Psychologists. I do not know, although I think that there have to be several different systems. They need to be interconnected in some way, and I do not know how.
3. Short Term Sensory Store (STSS): Sensory memory, generally very short term (250 ms-2 sec.) Often thought of as pre-attentional. Agent must attend to components of this memory in order for it to have any permanence.
3a. Visual sensory memory investigated by George Sperling (1960). One has a representation of a 3 X 4  array of letters for a brief period of time. It fades within approximately 1 sec. or it can be displaced with new visual input.
3b. Auditory sensory memory: One can sometimes attend to an auditory stimulus after it was heard preattentively (Based primarily on anecdotes).
Flashbulb images: long lasting (several seconds) positive sensory images are formed when there is a sudden flash onto a relatively dark adapted eye.
4. Short term store (STS or STM) Information that is attended to enters the STS. This is presumed to be a limited capacity system containing either 7 +/- 2 chunks or fewer depending on different experimental techniques. George Miller (1956) invented the concept of chunk and discussed the magic number 7 plus or minus 2. Memory span tends to lead to a 7 chunk memory; but recency effects tend to give a much smaller (2 or 3) STS size.
4a. Phonological loop.--Baddeley, How many items in a new list you can recall is a function of how long each of the items is and how long it takes to say the list. The longer it takes to say it, the fewer members one can recall.
4b. Brown and Peterson method for studying STS.  Present a three unit item to recall. Give the subject a task that involves his/her attention for differing amounts of time. Then recall the item. After only a few seconds it is very difficult to do.
4c. Proactive interference in the Brown-Peterson task. The more times you have done this task, the greater the forgetting. These results show that the associations that one has are not simply overridden by the new associations in short term memory tasks.
5. Release from proaction" (C. Wickens 19??) that shows that if the category from which the stimuli come is changed performance improves. Categories and proaction. This shows that short term memory tasks are not immune to meaning manipulations.
6. Working or primary memory--When people need to use their memory for certain tasks, the memory becomes much more accessible than it had been before. We often say that certain parts of the memory has become activated. An agent can make use of certain information much faster than s/he could without the activation. This active memory seems to have many of the same features as memory for information which has just been acquired and is currently in STS. The overlap is so great, it seems to me that working memory and STS when in the verbal domain are the same. The difference being the source of the information into the memory. If it comes primarily from external stimulation it's STS and if it's recalled from LTS it's working memory.
Baddeley actually suggests that there are different working memory workstations in the head, depending on what the particular task is. Two he identified are the Phonological loop, mentioned earlier, and the Visuospatial sketchpad to manipulate visual memories. I would expect him to suggest others for non-visuospatial, non verbal tasks.
7. Delayed response: A way of measuring STS in animals. An animal is shown a reward that is then hidden under an object. After a period of time the animal is released to get the reward. There are  evolutionary components to such tasks, higher mammals are more successful than lower ones. rat, raccoon, monkey, chimpanzee. Different tasks may have different brain loci dependent upon task and content.
II. Structure of Memory
1. Associative priming. Lexical decision task. Meyer and Schveneveldt (1971). Given a string of letters, answer as quickly as possible whether it is a word or not. If a word follows a word it is associated with RT is faster. Items in memory are associated with one another.
2.  Free recall Bousfield (1953) and Category clustering. If one is given a random list of words to recall in any order, the words will not be recalled randomly. If sprinkled within the list are words that are members of different categories, the members of the same category will tend to be recalled together. This is true for words that are associated with one another (Jenkins, 1952), and if there are no obvious relations to one another a subject will impart a structure to it. Tulving (1962) called this subjective organization.
3. False memories The method was first used by Deese (1959) and has been extended more recently by Roediger and McDermott (1995). The task is like many other memory experiments. A sequence of words is presented (verbally or visually) and the observer is to subsequently classify a set of words as either in the sequence (old), or not in the sequence (new). What differentiates this experiment from other memory experiments is that the sequences are specially designed to bias observers to report a particular word that was not included in the list. When people report that one of these words was in the sequence, but it really was not, they are having a false memory. In some cases people will report that they vividly report seeing (or hearing) the word, so their memory is very strong, despite its inaccuracy.

There are many studies, some done in the early part of the 20th century that show that people often recall things that have not actually occurred.
Many of these are related to making sense out of the information

• Brewer (1977) found that people often make pragmatic inferences in remembering sentences. People will remember the implications of a sentence, sometimes with more certainty than the sentence itself.
• Loftus found that misleading questions about some event leads subjects to select erroneous choices as being the ones that they had seen over correct ones. These often occur when the question deliberately misrepresents presupposed information.
• Pompi and Lachman. The meaning of a discourse controls the words that are "remembered" more than the associations of the pords presented. This leads me to believe that the associative networks in the Deese method is primarily a semantic field network. When some discourse is heard or read, people try to make sense out of it and they often misrepreseent the details of what they have heard. [see demonstration].  Often, if a subject is given a list to recall that has some structure to it, or if there are other reasons to not get it all, as was demonstrated when we saw that meaning was retained even though the specifics of the stimulus is not recalled.

Implicit memory tasks are those that show differences in performance as a function of what was remembered.  These two memory systems seem to have different parameters that effect them. Explicit memory is best with deeper and more meaningful and elaborate situations. Implicit often is more superficial and tied more closely with the stimulus situation.

• Seeing whether they can recognize a word when it is flashed very briefly.
• Recognizing a word that has some of the information in the stimulus obliterated.
• Generating a word under different conditions, e.g. to an associate of that and other words.
• Finding a word when given partial information; e.g. word completion task.
• Solving a mechanical puzzle;
• There is some evidence that implicit memory can take place while a patient is under general anesthesia and is supposedly "out". Subjects heard lists of words while under anesthesia for surgery and then were given a word completion task or a word recognition task. (Bonebakker, et al. (1996) Memory & Cognition, 24, 766-776.

to Chapter 7:

There are interesting demonstrations on the Purdue Cog Lab http://coglab.psych.purdue.edu/coglab/
 Peterson and Peterson,. Deese or Roediger and McClelland's False memory stuff, Sternberg's interrogating STM, Memory Span, Serial Position, and Sperling type STSS.